ICAART 2011 Abstracts

Area 1 - Artificial Intelligence
Full Papers Short Papers
Area 2 - Agents
Full Papers Short Papers

Area 1 - Artificial Intelligence

Full Papers
Paper Nr: 30
Title:

A RELATIONSHIP BETWEEN CROSS-VALIDATION AND VAPNIK BOUNDS ON GENERALIZATION OF LEARNING MACHINES
Authors:

Przemysław Klęsk
Abstract: Typically, the n-fold cross-validation is used both to: (1) estimate the generalization properties of a model of fixed complexity, (2) choose from a family of models of different complexities, the one with the best complexity, given a data set of certain size. Obviously, it is a time-consuming procedure. A different approach — the Structural Risk Minimization is based on generalization bounds of learning machines given by Vapnik (Vapnik, 1995a; Vapnik, 1995b). Roughly speaking, SRM is O(n) times faster than n-fold cross-validation but less accurate. We state and prove theorems, which show the probabilistic relationship between the two approaches. In particular, we show what e-difference between the two, one may expect without actually performing the crossvalidation. We conclude the paper with results of experiments confronting the probabilistic bounds we derived.
Download
Paper Nr: 32
Title:

UNRAVELLING BUENO DE MESQUITA’S GROUP DECISION MODEL

Authors:

Jason B. Scholz, Gregory J. Calbert and Glen A. Smith
Abstract: The development of societies of human and machine agents should benefit from an understanding of human group decision processes. Political Scientist and Professor, Bruce Bueno De Mesquita has made significant claims for the predictive accuracy of his computational model of group decision making, receiving much popular press including newspaper articles, books and a television documentary entitled “The New Nostradamus”. Despite these and many journal and conference publications related to the topic, no clear elicitation of the model exists in the open literature. We expose and present the model by careful navigation of the literature and illustrate the soundness of our interpretation by replicating De Mesquita’s own results. We also discuss concerns regarding model sensitivity and convergence.
Download
Paper Nr: 49
Title:

A DYNAMICAL MODEL FOR SIMULATING A DEBATE OUTCOME

Authors:

A. Imoussaten, J. Montmain, A. Rico and F. Rico
Abstract: A group of agents is faced with collective decisional problems. The corresponding debate is seen as a dynamical process. A first theoretical model based upon a muticriteria decision framework was proposed in (Rico et al., 2004) but without semantic justifications and explicit dynamical representation. A second descriptive model was proposed in (Imoussaten et al., 2009) where social influences and argumentation strategy govern the dynamics of the debate. This paper aims at justifying the equations introduced in (Rico et al., 2004) with the semantics concepts reported in (Imoussaten et al., 2009) to provide a model of a debate in the framework of control theory that explicitly exhibits dynamical aspects and offers further perspectives for control purposes of the debate.
Download
Paper Nr: 54
Title:

OPTIMAL SAMPLE SELECTION FOR BATCH-MODE REINFORCEMENT LEARNING
Authors:

Emmanuel Rachelson, François Schnitzler, Louis Wehenkel and Damien Ernst
Abstract: We introduce the Optimal Sample Selection (OSS) meta-algorithm for solving discrete-time Optimal Control problems. This meta- algorithm maps the problem of finding a near-optimal closed-loop policy to the identification of a small set of one-step system transitions, leading to high-quality policies when used as input of a batch-mode Reinforcement Learning (RL) algorithm. We detail a particular instance of this OSS meta-algorithm that uses tree-based Fitted Q-Iteration as a batch-mode RL algorithm and Cross Entropy search as a method for navigating efficiently in the space of sample sets. The results show that this particular instance of OSS algorithms is able to identify rapidly small sample sets leading to high-quality policies.
Download
Paper Nr: 57
Title:

EXTRACTION OF FUNCTION FEATURES FOR AN AUTOMATIC CONFIGURATION OF PARTICLE SWARM OPTIMIZATION
Authors:

Tjorben Bogon, Georgios Poursanidis, Andreas D. Lattner and Ingo J. Timm
Abstract: In this paper we introduce a new approach for automatic parameter configuration of Particle Swarm Optimization (PSO) by using features of objective function evaluations for classification. This classification utilizes a decision tree that is trained by using 32 function features. To classify different functions we compute features of the function from observed PSO behavior. These features are an adequate description to compare different objective functions. This approach leads to a trained classifier which gets as input a function and returns a parameter set. Using this parameter set leads to an equal or better optimization process compared to the standard parameter settings of Particle Swarm Optimization on selected test functions.
Download
Paper Nr: 59
Title:

PARTIALLY-CONTROLLED MARKOV DECISION PROCESSES FOR COLLISION AVOIDANCE SYSTEMS
Authors:

Mykel J. Kochenderfer and James P. Chryssanthacopoulos
Abstract: Deciding when and how to avoid collision in stochastic environments requires accounting for the likelihood and relative costs of future sequences of outcomes in response to different sequences of actions. Prior work has investigated formulating the problem as a Markov decision process, discretizing the state space, and solving for the optimal strategy using dynamic programming. Experiments have shown that such an approach can be very effective, but scaling to higher-dimensional problems can be challenging due to the exponential growth of the discrete state space. This paper presents an approach that can greatly reduce the complexity of computing the optimal strategy in problems where only some of the dimensions of the problem are controllable. The approach is demonstrated on an airborne collision avoidance problem where the system must recommend maneuvers to an imperfect pilot.
Download
Paper Nr: 63
Title:

A BAYESIAN METHOD FOR THE DETECTION OF EPISTASIS IN QUANTITATIVE TRAIT LOCI USING MARKOV CHAIN MONTE CARLO MODEL COMPOSITION WITH RESTRICTED MODEL SPACES
Authors:

Edward L. Boone, Susan J. Simmons and Karl Ricanek
Abstract: Epistasis or the interaction between loci on a genome is of great interest to geneticists. Herein, a powerful Bayesian method utilizing Markov chain Monte Carlo model composition approach using restricted spaces is developed for identifying epistatic effects in Recombinant Inbred Lines (RIL). The method is verified through a simulation study and applied to an Arabidopsis thaliana data set with cotyledon as the quantitative trait.
Download
Paper Nr: 78
Title:

INTEREST-BASED PREFERENCE REASONING
Authors:

Wietske Visser, Koen V. Hindriks and Catholijn M. Jonker
Abstract: In the context of practical reasoning, such as decision making and negotiation, it is necessary to model preferences over possible outcomes. Such preferences usually depend on multiple criteria. We argue that the criteria by which outcomes are evaluated should be the satisfaction of a person’s underlying interests: the more an outcome satisfies his interests, the more preferred it is. Underlying interests can explain and eliminate conditional preferences. Also, modelling interests will create a better model of human preferences, and can lead to better, more creative deals in negotiation. We present an argumentation framework for reasoning about interest-based preferences. We take a qualitative approach and provide the means to derive both ceteris paribus and lexicographic preferences.
Download
Paper Nr: 86
Title:

CLUSTERING WITH GRANULAR INFORMATION PROCESSING
Authors:

Urszula Kużelewska
Abstract: Clustering is a part of data mining domain. Its task is to detect groups of similar objects on the basis of established similarity criterion. Granular computing (GrC) includes methods from various areas with the aim to support human with better understanding analyzed problem and generated results. Granular computing techniques create and/or process data portions named as granules identified with regard to similar description, functionality or behavior. Interesting characteristic of granular computation is offer of multi-perspective view of data depending on required resolution level. Data granules identified on different levels of resolution form a hierarchical structure expressing relations between objects of data. A method proposed in this article performs creation data granules by clustering data in form of hyperboxes. The results are compared with clustering of point-type data with regard to complexity, quality and interpretability.
Download
Paper Nr: 154
Title:

PROVIDING DELIBERATION TO EMOTIONAL AGENTS
Authors:

Daniel Pérez-Pinillos, Susana Fernández and Daniel Borrajo
Abstract: Modelling real persons or virtual agents motivations, personality and emotions is a key feature of many user-oriented applications. Most of the previous work has defined rich cognitive models of motivations, personality and emotions, but have relied on some kind of reactive scheme of problem solving and execution. Instead, this work proposes a deliberative emotional model for virtual agents based in their basic needs, preferences and personality traits. More specifically, we advocate the integration of these comprehensive agents models within deliberative automated planning techniques, so that plans to be executed by agents to achieve their goals already incorporate reasoning at the emotional level.
Download
Paper Nr: 188
Title:

EFFICIENT SYMBOLIC SUPERVISORY SYNTHESIS AND GUARD GENERATION - Evaluating Partitioning Techniques for the State-space Exploration
Authors:

Z. Fei, S. Miremadi, K. Åkesson and B. Lennartson
Abstract: The supervisory control theory (SCT) is a model-based framework which automatically synthesizes a supervisor that restricts a plant to be controlled based on specifications to be fulfilled. Two main problems typically encountered in industrial applications prevent SCT from having a major breakthrough. First, the supervisor which is synthesized automatically from the given plant and specification models might be incomprehensible to the users. To tackle this problem, in our previous work, an approach was presented to extract compact propositional formulae from the supervisor represented symbolically by binary decision diagrams (BDD) and attach them to the original models. However, this approach, which computes the supervisor symbolically in the conjunctive way, might lead to another problem: the state-space explosion, because of the large number of intermediate BDD nodes during computation. In this paper, we introduce an alternative approach to alleviate this problem, which is based on the disjunctive partitioning technique including a set of selection heuristics. Then this approach is adapted to the guard generation procedure. Finally, the efficiency of the presented approach is demonstrated on a set of benchmark examples.
Download
Paper Nr: 197
Title:

NBDI: AN ARCHITECTURE FOR GOAL-ORIENTED NORMATIVE AGENTS
Authors:

Baldoino F. dos S. Neto, Viviane Torres da Silva and Carlos J. P. de Lucena
Abstract: Normative regulation is a mechanism of social control that regulates the behaviour of autonomous agents indicating that such agents are permitted, obligated or prohibited to achieve something in the world, and defining rewards by stating stimulus to their fulfilment and punishments by discouraging their violation. Since goal-oriented agents‘ priority is the accomplishment of their own desires, they must evaluate the pros and cons associated with the fulfilment or violation of the norms before choosing to comply or not with them. In this context, we propose an extension to the BDI architecture by including norms related functions to check the incoming perceptions (including norms), select the norms they intend to fulfil based on the benefits they provide to the achievement of the agent‘s desires and intentions, and decide to cope or not with the norms while dropping, retaining or adopting new intentions. We demonstrate the applicability of our approach through an non- combatant evacuation scenario implemented by using the Normative Jason platform.
Download
Paper Nr: 217
Title:

EVALUATING THE COST OF SUPPORTING INTERACTION AND SIMULATION THROUGH THE ENVIRONMENT
Authors:

Flavien Balbo, Fabien Badeig and Julien Saunier
Abstract: The environment has emerged as a powerful first-order abstraction in Multi-Agent Systems (MAS), as well as a critical building block. One benefit is to reduce the complexity of the agents by delegating to the environment a part of the tasks of the system. This delegating process provides a flexible way to exchange information and to coordinate the agents thanks to the environment. The counterpart is a centralization of a part of the MAS processes inside the environment. In this paper, we present the modeling of an environment for multi-agent communication and simulation. Our proposition enables the addition of advanced features to the MAS like multi-party communications (communication) and contextual activation (simulation). We evaluate the cost of this environment process and compare it to the execution of the same tasks in the agents for communication and simulation.
Download
Paper Nr: 229
Title:

AN ADAPTIVE SELECTIVE ENSEMBLE FOR DATA STREAMS CLASSIFICATION
Authors:

Valerio Grossi and Franco Turini
Abstract: The large diffusion of different technologies related to web applications, sensor networks and ubiquitous computing, has introduced new important challenges for the data mining community. The rising need of analyzing data streams introduces several requirements and constraints for a mining system. This paper analyses a set of requirements related to the data streams environment, and proposes a new adaptive method for data streams classification. The system employs data aggregation techniques that, coupled with a selective ensemble approach, perform the classification task. The approach adopts the behaviour of the selective ensemble by dynamically updating the threshold for enabling the classifiers. The system is explicitly conceived to satisfy these requirements even in the presence of concept drifting.
Download
Paper Nr: 242
Title:

SOLVING NON BINARY CONSTRAINT SATISFACTION PROBLEMS WITH DUAL BACKTRACKING ON HYPERTREE DECOMPOSITION
Authors:

Zineb Habbas, Kamal Amroun and Daniel Singer
Abstract: Solving a CSP (Constraint Satisfaction Problem) is NP-Complete in general. However, there are various classes of CSPs that can be solved in polynomial time. Some of them can be identified by analyzing their structure. It is theoretically well established that a tree (or hypertree) structured CSP can be solved in a backtrack-free way leading to tractability. Different methods exist for converting CSPs in a tree (or hypertree) structured representation. Among these methods Hypertree Decomposition has been proved to be the most general one for non-binary CSPs. Unfortunately, in spite of its good theoretical bound, the unique algorithm for solving CSP from its hypertree structure is inefficient in practice due to its memory explosion. To overcome this problem, we propose in this paper a new approach exploiting a Generalized Hypertree Decomposition. We present the so called HD DBT algorithm (Dual BackTracking algorithm guided by an order induced by a generalized Hypertree Decomposition). Different heuristics and implementations are presented showing its practical interest.
Download
Paper Nr: 247
Title:

LOCAL FEATURE BASED IMAGE SIMILARITY FUNCTIONS FOR KNN CLASSIFICATION
Authors:

Giuseppe Amato and Fabrizio Falchi
Abstract: In this paper we consider the problem of image content recognition and we address it by using local features and kNN based classification strategies. Specifically, we define a number of image similarity functions relying on local feature similarity and matching with and without geometric constrains. We compare their performance when used with a kNN classifier. Finally we compare everything with a new kNN based classification strategy that makes direct use of similarity between local features rather than similarity between entire images. As expected, the use of geometric information offers an improvement over the use of pure image similarity. However, surprisingly, the kNN classifier that use local feature similarity has a better performance than the others, even without the use of geometric information. We perform our experiments solving the task of recognizing landmarks in photos.
Download
Paper Nr: 267
Title:

TRANSLATING WEB SERVICES COMPOSITION PLANS TO OWL-S DESCRIPTIONS

Authors:

Eva Ziaka, Dimitris Vrakas and Nick Bassiliades
Abstract: Web Services technology has led to simpler and more rapid development of Web Applications with improved functionality by which several platforms through the globe can communicate to exchange data and cooperate for problem solving. Methods for automated web services composition are studied so as to enhance this type of software development. Many studies focus on converting the composition problem to a planning problem and solving it using known planning algorithms. This paper suggests a method for translating the produced PDDL plans of the above algorithms to OWL-S descriptions of the final composite web services. The result is a totally new web service that can later be discovered and invoked or even take part in a new composition.
Download
Paper Nr: 272
Title:

AN APPROACH TO SIGNIFICANCE ESTIMATION FOR SIMULATION STUDIES
Authors:

Andreas D. Lattner, Tjorben Bogon and Ingo J. Timm
Abstract: Simulation is widely used in order to evaluate system changes, to perform parameter optimization of systems, or to compare existing alternatives. Assistance systems for simulation studies can support the user by performing monotonous tasks and keeping track of relevant results. In this paper we present an approach to significance estimation in order to estimate, if – and when – statistically significant results are expected for certain investigations. This can be used for controlling simulation runs or providing information to the user for interaction. We introduce two approaches: one for the classification if significance is expected to occur for given samples and another for the prediction of needed replications until significance migh
Download
Short Papers
Paper Nr: 24
Title:

A RECOMMENDATION ALGORITHM FOR PERSONALIZED ONLINE NEWS BASED ON COLLECTIVE INTELLIGENCE AND CONTENT
Authors:

Giovanni Giuffrida and Calogero G. Zarba
Abstract: We present a recommendation algorithm for online news based on collective intelligence and content. When a user asks for personalized news, our algorithm recommends news articles that (i) are popular among the members of the online community (the collective intelligence part), and (ii) are similar in content to the news articles the user has read in the past (the content part). Our algorithm computes its recomendations based on the collective behavior of the online users, as well as on the feedback the users provide to the algorithm’s recommendations. The users’ feedback can moreover be used to measure the effectiveness of our recomendation algorithm in terms of the information retrieval concepts of precision and recall. The cornerstone of our recommendation algorithm is a basic relevance algorithm that computes how relevant a news article is to a given user. This basic relevance algorithm can be optimized in order to obtain a faster online response at the cost of minimal offline computations. Moreover, it can be turned into an approximated algorithm for an even faster online response.
Download
Paper Nr: 36
Title:

QUORUM SENSING FOR COLLECTIVE ACTION AND DECISION-MAKING IN MOBILE AUTONOMOUS TEAMS

Authors:

Donald A. Sofge and William F. Lawless
Abstract: Design of controllers for teams of mobile autonomous systems presents many challenges that have been addressed in biological systems, such as behavior-based control paradigms that are decentralized, distributed, scalable, and robust. Quorum sensing is a distributed, decentralized decision-making process used by bacteria and by social insects to coordinate group behaviors and perform complex tasks. It is used by bacteria to control the colony behavior for a variety of functions, such as biofilm construction or initiating pathogenicity inside a host. It is used by social insects including the ant Temnothorax albipennis to collectively evaluate and select from amongst potentially many new nesting sites.Honeybees (Apis mellifera) use quorum sensing to collectively choose a new nesting site when the swarm grows too large and needs to split. It is shown that the quorum sensing paradigm may be used to provide robust decentralized team coordination and collective decision-making in mobile autonomous teams performing complex tasks. In this effort quorum sensing-inspired techniques are developed and applied to the design of a decentralized controller for a team of mobile autonomous agents surveying a field containing buried landmines.
Download
Paper Nr: 47
Title:

MOLECULAR FUZZY INFERENCE ENGINES - Development of Chemical Systems to Process Fuzzy Logic at the Molecular Level
Authors:

Pier Luigi Gentili
Abstract: Current Information Technology is pursuing a revolution in the design of computing machines: it is trying to pass from macroscopic processors miniaturized through top-down approaches, to microscopic processors made of single molecules assembled through bottom-up approaches. When computations are carried out by single atoms and molecules, quantum logic can be processed. It is difficult to devise a quantum computer due to the decoherent effects exerted by the surrounding environment. However, it is still possible to work out with molecules, by abandoning the lure of quantum logic and processing classical logic. Single molecules make binary computations, whereas ensembles of molecules can be used to implement either Boolean logic gates or Fuzzy inference engines. The behaviours of two chemical compounds after photo-excitation are described as examples of quantum systems whereby Fuzzy logic can be processed by exploiting the decoherent effects exerted by the surrounding microenvironment.
Download
Paper Nr: 53
Title:

DESIGN OF AUTOMATICALLY ADAPTABLE WEB WRAPPERS
Authors:

Emilio Ferrara and Robert Baumgartner
Abstract: Nowadays, the huge amount of information distributed through the Web motivates studying techniques to be adopted in order to extract relevant data in an efficient and reliable way. Both academia and enterprises developed several approaches of Web data extraction, for example using techniques of artificial intelligence or machine learning. Some commonly adopted procedures, namely wrappers, ensure a high degree of precision of information extracted from Web pages, and, at the same time, have to prove robustness in order not to compromise quality and reliability of data themselves. In this paper we focus on some experimental aspects related to the robustness of the data extraction process and the possibility of automatically adapting wrappers. We discuss the implementation of algorithms for finding similarities between two different version of a Web page, in order to handle modifications, avoiding the failure of data extraction tasks and ensuring reliability of information extracted. Our purpose is to evaluate performances, advantages and draw-backs of our novel system of automatic wrapper adaptation.
Download
Paper Nr: 68
Title:

EVALUATION OF A USER-ADAPTED SPOKEN LANGUAGE DIALOGUE SYSTEM - Measuring the Relevance of the Contextual Information Sources
Authors:

Juan Manuel Lucas-Cuesta, Fernando Fernández-Martínez, G. Dragos Rada, Syaheerah L. Lutfi and Javier Ferreiros
Abstract: We present an evaluation of a spoken language dialogue system with a module for the management of user-related information, stored as user preferences and privileges. The flexibility of our dialogue management approach, based on Bayesian Networks (BN), together with a contextual information module, which performs different strategies for handling such information, allows us to include user information as a new level into the Context Manager hierarchy. We propose a set of objective and subjective metrics to measure the relevance of the different contextual information sources. The analysis of our evaluation scenarios shows that the relevance of the short-term information (i.e. the system status) remains pretty stable throughout the dialogue, whereas the dialogue history and the user profile (i.e. the middle-term and the long-term information, respectively) play a complementary role, evolving their usefulness as the dialogue evolves.
Download
Paper Nr: 81
Title:

IT’S ANSWER TIME - Taking the Next Step in Question-Answering
Authors:

Ana Cristina Mendes and Luísa Coheur
Abstract: After all the work done in tasks like question classification, query expansion or information extraction in QA, we consider that some efforts should now be put specially on giving the answer to the user. In this paper we adopt the concept of cooperative answer – that is, a correct, useful and non-misleading answer – since it is our opinion that finding and presenting the cooperative answer to the user is one of the next challenges in QA.With that goal in mind, we focus on three main aspects that should deserve the attention of the QA community: the ability of systems to relate the candidate answers for a question; their ability to decide which candidates are possible final answers, given the question, but also the user who posed it; and, finally, the ability of generating the final answer in a cooperative way.
Download
Paper Nr: 90
Title:

FEATURE INDUCTION OF LINEAR-CHAIN CONDITIONAL RANDOM FIELDS - A Study based on a Simulation
Authors:

Dapeng Zhang and Bernhard Nebel
Abstract: Conditional Random Fields (CRFs) is a probabilistic framework for labeling sequential data. Several approaches were developed to automatically induce features for CRFs. They have been successfully applied in real-world applications, e.g. in natural language processing. The work described in this paper was originally motivated by processing the sequence data of table soccer games. As labeling such data is very time consuming, we developed a sequence generator (simulation), which creates an extra phase to explore several basic issues of the feature induction of linear-chain CRFs. First, we generated data sets with different configurations of overlapped and conjunct atomic features, and discussed how these factors affect the induction. Then, a reduction step was integrated into the induction which maintained the prediction accuracy and saved the computational power. Finally, we developed an approach which consists of a queue of CRFs. The experiments show that the CRF queue achieves better results on the data sets in all the configurations.
Download
Paper Nr: 92
Title:

MULTIOBJECTIVE OPTIMIZATION OF THE 3D TOPOLOGICAL ACTIVE VOLUME SEGMENTATION MODEL
Authors:

Jorge Novo, Manuel G. Penedo and José Santos
Abstract: In this work it is proposed an evolutionary multiobjective methodology for the optimization of topological active volumes. This is a 3D deformable model that integrates features of region-based and boundary-based segmentation techniques. The model deformation is controlled by energy functions that must be minimized. Most optimization algorithms need an experimental tuning of the energy parameters of the model in order to obtain the best adjusted segmentation. To avoid the step of the parameter tuning, we developed an evolutionary multiobjective optimization that considers the optimization of several objectives in parallel. The proposed methodology is based on the SPEA2 algorithm, adapted to our application, to obtain the Pareto optimal individuals. The proposed method was tested on several representative images from different domains yielding highly accurate results.
Download
Paper Nr: 114
Title:

EPISTEMIC REASONING FOR AMBIENT INTELLIGENCE
Authors:

Theodore Patkos and Dimitris Plexousakis
Abstract: Ambient Intelligence is an emerging discipline that requires the integration of expertise from a multitude of scientific fields. The role of Artificial Intelligence is crucial not only for bringing intelligence to everyday environments, but also for providing the means for the different disciplines to collaborate. In this paper we highlight the importance of reasoning under partial observability in such dynamic and context-rich domains and illustrate the integration of an epistemic theory to an operational Ambient Intelligence infrastructure.
Download
Paper Nr: 115
Title:

AUTOMATIC STATE SPACE AGGREGATION USING A DENSITY BASED TECHNIQUE
Authors:

Steven Loscalzo and Robert Wright
Abstract: Applying reinforcement learning techniques in continuous environments is challenging because there are infinitely many states to visit in order to learn an optimal policy. To make this situation tractable, abstractions are often used to reduce the infinite state space down to a small and finite one. Some of the more powerful and commonplace abstractions, tiling abstractions such as CMAC, work by aggregating many base states into a single abstract state. Unfortunately, significant manual effort is often necessary in order to apply them to nontrivial control problems. Here we develop an automatic state space aggregation algorithm, Maximum Density Separation, which can produce a meaningful abstraction with minimal manual effort. This method leverages the density of observations in the space to construct a partition and aggregate states in a dense region to the same abstract state. We show that the abstractions produced by this method on two benchmark reinforcement learning problems can outperform fixed tiling methods in terms of both the convergence rate of a learning algorithm and the number of abstract states needed.
Download
Paper Nr: 120
Title:

PRE-PROCESSING TASKS FOR RULE-BASED ENGLISH-KOREAN MACHINE TRANSLATION SYSTEM

Authors:

Sung-Dong Kim
Abstract: This paper presents necessary pre-processing tasks for practical English-Korean machine translation. The pre-processing task consists of a problem that requires pre-processing and a solution for the problem. There are many differences between English and Korean, so it is difficult to resolve the differences using parsing and transfer rules. Also, source sentences often include non-word elements, such as parentheses, quotation marks, and list markers. In order to resolve the differences efficiently and make source sentences appropriate to translation system by arranging them, we propose pre-processing for source sentences. This paper studies various pre-processing tasks and classifies into several groups according to the time when the tasks are performed in English-Korean machine translation system. In experiment, we show the usefulness of the defined pre-processing tasks for generating better translation results.
Download
Paper Nr: 125
Title:

EMBEDDED INCREMENTAL FEATURE SELECTION FOR REINFORCEMENT LEARNING
Authors:

Robert Wright, Steven Loscalzo and Lei Yu
Abstract: Classical reinforcement learning techniques become impractical in domains with large complex state spaces. The size of a domain’s state space is dominated by the number of features used to describe the state. Fortunately, in many real-world environments learning an effective policy does not usually require all the provided features. In this paper we present a feature selection algorithm for reinforcement learning called Incremental Feature Selection Embedded in NEAT (IFSE-NEAT) that incorporates sequential forward search into neuroevolutionary algorithm NEAT. We provide an empirical analysis on a realistic simulated domain with many irrelevant and relevant features. Our results demonstrate that IFSE-NEAT selects smaller and more effective feature sets than alternative approaches, NEAT and FS-NEAT, and superior performance characteristics as the number of available features increases.
Download
Paper Nr: 134
Title:

FALSE ALARM FILTERING IN A VISION TRAFFIC SIGN RECOGNITION SYSTEM - An Approach based on AdaBoost and Heterogeneity of Texture
Authors:

Sergio Lafuente-Arroyo, Saturnino Maldonado-Bascón, Hilario Gómez-Moreno and Pedro Gil-Jiménez
Abstract: The high variability of road sign appearance and the variety of different classes have made the recognition of pictograms a high computational load problem in traffic sign detection based on computer vision. In this paper false alarms are reduced significantly by designing a cascade filter based on boosting detectors and a generative classifier based on heterogeneity of texture. The false alarm filter allows us to discard many false positives using a reduced selection of features, which are chosen from a wide set of features. Filtering is defined as a binary problem, where all speed limit signs are grouped together against noisy examples and it is the previous stage to the input of a recognition module based on Support Vector Machines (SVMs). In a traffic sign recognition system, the number of candidate blobs detected is, in general, much higher than the number of traffic signs. As asymmetry is an inherent problem, we apply a different treatment for false negatives (FN) and false positives (FP). The global filter offers high accuracy. It achieves very low false alarm ratio with low computational complexity.
Download
Paper Nr: 136
Title:

MULTI-AGENT SOFT CONSTRAINT AGGREGATION - A Sequential Approach
Authors:

Giorgio Dalla Pozza, Francesca Rossi and K. Brent Venable
Abstract: We consider a scenario where several agents express their preferences over a common set of variable assignments, by means of a soft constraint problem for each agent, and we propose a procedure to compute a variable assignment which satisfies the agents’ preferences at best. Such a procedure considers one variable at a time and, at each step, asks all agents to express its preferences over the domain of that variable. Based on such preferences, a voting rule is used to decide on which value is the best for that variable. At the end, the values chosen constitute the returned variable assignment. We study several properties of this procedure and we show that the use of soft constraints allows for a great flexibility on the preferences of the agents, compared to similar work in setting where agents model their preferences via CP-nets, where several restrictions on the agents’ preferences need to be imposed to obtain similar properties.
Download
Paper Nr: 153
Title:

SEMANTIC CLUSTERING BASED ON ONTOLOGIES - An Application to the Study of Visitors in a Natural Reserve
Authors:

Montserrat Batet, Aïda Valls and Karina Gibert
Abstract: The development of large ontologies for general and specific domains provides new tools to improve the quality of data mining techniques such as clustering. In this paper we explain how to improve clustering results by exploiting the semantics of categorical data by means of ontologies and how this semantics can be included into a hierarchical clustering method. We want to prove that when the conceptual meaning of the values is taken into account, it is possible to find a better interpretation of the clusters. This is demonstrated with the analysis of real data collected from visitors to of a Natural Reserve. The results of our methodology are compared with the ones obtained with a classical multivariate analysis done in the same database.
Download
Paper Nr: 176
Title:

SVM-BASED PARAMETER SETTING OF SELF-QUOTIENT e-FILTER AND ITS APPLICATION TO NOISE ROBUST HUMAN DETECTION
Authors:

Mitsuharu Matsumoto
Abstract: This paper describes SVM-based parameter setting of self-quotient ɛ-filter (SQEF), and its application to noise robust human detection combining SQEF, histograms of oriented gradients (HOG), and support vector machine (SVM). Although human detection combining HOG and SVM is a powerful approach, as it uses local intensity gradients, it is difficult to handle noise corrupted images. On the other hand, although human detection combining SQEF, HOG and SVM can realize noise robust human detection, SQEF requires manual parameter setting. Our aim is not only to train SVM but also to adjust the parameter of self-quotient ɛ-filter using the trained SVM in training procedure. The experimental results show that we can realize noise robust human detection by using SQEF with the obtained parameter, HOG and SVM trained by intact images without noise.
Download
Paper Nr: 179
Title:

THE MULTI-AGENT PLANNING PROBLEM
Authors:

Tamás Kalmár-Nagy and Giovanni Giardini
Abstract: The purpose of this paper is to present a Multi-Agent planner for a team of autonomous agents. The approach is demonstrated by the Multi-Agent Planning Problem, which is a variant of the classical Multiple Traveling Salesmen Problem (MTSP): given a set of n goals/targets and a team of m agents, the optimal team strategy consists of finding m tours such that each target is visited only once and by only one agent, and the total cost of visiting all nodes is minimal. The proposed solution method is a Genetic Algorithm Inspired Steepest Descent (GAISD) method. To validate the approach, the method has been benchmarked against MTSPs and routing problems. Numerical experiments demonstrate the goodness of the approach.
Download
Paper Nr: 180
Title:

A SPATIAL QUERY LANGUAGE FOR PRESENTATION-ORIENTED DOCUMENTS
Authors:

Ermelinda Oro, Francesco Riccetti and Massimo Ruffolo
Abstract: In last years the huge relevance of accessing and acquiring information made available byWeb (HTML) pages and business (PDF) documents has grown much further. In this paper we present a textual query language, named ViQueL, whose main feature is to identify and extract relevant information from HTML and PDF documents on the base of their visual appearance by using easy-to-write queries. The proposed language is founded on spatial grammars, i.e. context free grammars extended by spatial constructs. Despite a considerable expressive power, combined complexity of ViQueL is in P-Time. Moreover, experiments show that ViQueL is reasonably efficient for real-life extraction tasks.
Download
Paper Nr: 187
Title:

TOWARDS ROUTING FOR AUTONOMOUS ROBOTS - Using Constraint Programming in an Anytime Path Planner
Authors:

Roman Barták, Michal Zerola and Stanislav Slušný
Abstract: Path planning is one of the critical tasks for autonomous robots. In this paper we study the problem of finding the shortest path for a robot collecting waste spread over the area such that the robot has a limited capacity and hence during the route it must periodically visit depots/collectors to empty the collected waste. This is a variant of often overlooked vehicle routing problem with satellite facilities. We present two approaches for this optimisation problem both based on Constraint Programming techniques. The former one is inspired by the operations research model, namely by the network flows, while the second one is driven by the concept of finite state automaton. The experimental comparison and enhancements of both models are discussed with emphasis on the further adaptation to the real world environment.
Download
Paper Nr: 190
Title:

IMPROVING THE MAPPING PROCESS IN ONTOLOGY-BASED USER PROFILES FOR WEB PERSONALIZATION SYSTEMS
Authors:

Ahmad Hawalah and Maria Fasli
Abstract: Web personalization systems that have emerged in recent years enhance the retrieval process based on each user’s interests and preferences. A key feature in developing an effective web personalization system is to build and model user profiles accurately. In this paper, we propose an approach that implicitly tracks users’ browsing behaviour in order to build an ontology-based user profile. The main goal of this paper is to investigate techniques to improve the accuracy of this user profile. We focus in particular on the mapping process which involves mapping the collected web pages the user has visited to concepts in a reference ontology. For this purpose, we introduce two techniques to enhance the mapping process: one that maintains the user’s general and specific interests without the user’s involvement, and one that exploits browsing and search contexts. We evaluate the factors that impact the overall performance of both techniques and show that our techniques improve the overall accuracy of the user profile.
Download
Paper Nr: 198
Title:

ALL ABOUT MICROTEXT - A Working Definition and a Survey of Current Microtext Research within Artificial Intelligence and Natural Language Processing

Authors:

Jeffrey Ellen
Abstract: This paper defines a new term, ‘Microtext’, and takes a survey of the most recent and promising research that falls under this new definition. Microtext has three distinct attributes that differentiate it from the traditional free-text or unstructured text considered within the AI and NLP communities. Microtext is text that is generally very short in length, semi-structured, and characterized by amorphous or informal grammar and language. Examples of microtext include chatrooms (such as IM, XMPP, and IRC), SMS, voice transcriptions, and micro-blogging such as Twitter(tm). This paper expands on this definition, and provides some characterizations of typical microtext data. Microtext is becoming more prevalent. It is the thesis of this paper that the three distinct attributes of microtext yield different results and require different techniques than traditional AI and NLP techniques on long-form free text. By creating a working definition for microtext, providing a survey of the current state of research in the area, it is the goal of this paper to create an understanding of microtext within the AI and NLP communities.
Download
Paper Nr: 206
Title:

UNSUPERVISED ADAPTATION OF THE USER INTERESTS
Authors:

Lucas Marin, David Isern and Antonio Moreno
Abstract: One of the main problems in recommender systems is to ensure the quality of the user profile. This issue is particularly challenging if the user preferences may vary in time. This paper proposes a novel unsupervised algorithm to adapt dynamically the user profile, taking into account the interaction of the user with the system. The paper discusses the influence of the basic parameters of the adaptation algorithm and presents some promising preliminary results.
Download
Paper Nr: 212
Title:

AUGMENTED REALITY BASED INTELLIGENT INTERACTIVE E-LEARNING PLATFORM
Authors:

Héctor Martínez, Rafael del Hoyo, Luis Miguel Sanagustín, Isabelle Hupont, David Abadia and Carlos Sagüés
Abstract: e-Learning systems are continuously evolving in order to include new technologies that improve the education process. Some of the technologies that are being incorporated to the e-learning systems are related to virtual agents and Augmented Reality. The proposed architecture aims to offer a novel platform for non-programming experienced users to develop intelligent Augmented Reality e-learning applications by an intelligent fuzzy-rules-based framework. The applications consist of a series of interactive Augmented Reality exercises guided by an intelligent adaptive virtual tutor to help the student in the learning process.
Download
Paper Nr: 213
Title:

TEXT SEGMENTATION USING NAMED ENTITY RECOGNITION AND CO-REFERENCE RESOLUTION
Authors:

Pavlina Fragkou
Abstract: In this paper we examine the benefit of performing named entity recognition and co-reference resolution to a benchmark used for text segmentation. The aim here is to examine whether the incorporation of such information enhances the performance of text segmentation algorithms. The evaluation using three well known text segmentation algorithms leads to the conclusion that, the benefit highly depends on the segment's topic, the number of named entity instances appearing in it, as well as the segment's length.
Download
Paper Nr: 216
Title:

FORMALIZING DIALECTICAL REASONING FOR COMPROMISE-BASED JUSTIFICATION
Authors:

Hiroyuki Kido, Katsumi Nitta, Masahito Kurihara and Daisuke Katagami
Abstract: Chinese traditional philosophy regards dialectics as a style of reasoning that focuses on contradictions and how to resolve them, transcend them or find the truth in both. Compromise is considered to be one possible way to resolve conflicts dialectically. In this paper, we formalize dialectical reasoning as a way for deriving compromise. Both the definition of the notion of compromise and the algorithm for dialectical reasoning are proposed on an abstract complete lattice. We prove that the dialectical reasoning is sound and complete with respect to the compromise. We propose the concrete algorithm for dialectical reasoning characterized by definite clausal language and generalized subsumption. The algorithm is proved to be sound with respect to the compromise. Furthermore, we expand an argumentation system to handle compromise arguments, and illustrate that an agent bringing up a compromise argument realizes a compromise based justification towards argument-based deliberation.
Download
Paper Nr: 222
Title:

INTCARE - Multi-agent Approach for Real-time Intelligent Decision Support in Intensive Medicine

Authors:

Manuel Filipe Santos, Filipe Portela and Marta Vilas-Boas
Abstract: For an Intelligent Decision Support System to work in real-time, it is of great value the use of intelligent agents that cooperate with each other to accomplish their tasks. In a critical environment like an Intensive Care Unit, doctors should have the right information, at the right time, to better assist their patients. In this paper we present an architecture for a Multi-Agents System that will support doctors’ decision by in real-time, guaranteeing that all required clinical data is available and capable of predicting the patients’ condition for the next hour.
Download
Paper Nr: 223
Title:

EXPERIMENTS IN SHORT-TERM WIND POWER PREDICTION USING VARIABLE SELECTION
Authors:

Javier Lorenzo, Juan Méndez, Daniel Hernández and Modesto Castrillón
Abstract: In this paper some experiments have been realized to test how the introduction of variable selection has an effect on the predictor performance in short-term wind farm power prediction. Variable selection based on Kraskov estimation of the mutual information will be used due to its capability to deal with sets of continuous random variables. A Multilayer Percetron and a k-NN estimator will be the predictor based models with different topologies and number of neighbors. Experiments will be carried out with actual data of wind speed and power of an experimental wind farm. We also compute the output of an ideal wind turbine to enrich the dataset and estimate the effect of variable selection on one isolated turbine. This will allow us to define four different settings for the experiments which vary in the nature of the inputs to the model, wind speed, wind farm or isolated wind turbine power, and the predicted variable, wind farm or isolated wind turbine power.
Download
Paper Nr: 224
Title:

EVALUATING RERANKING METHODS USING WIKIPEDIA FEATURES
Authors:

Koji Kurakado, Tetsuya Oishi, Ryuzo Hasegawa, Hiroshi Fujita and Miyuki Koshimura
Abstract: Many people these days access a vast document on theWeb very often with the help of search engines such as Google. However, even if we use the search engine, it is often the case that we cannot find desired information easily. In this paper, we extract related words for the search query by analyzing link information and category structure. we aim to assist the user in retrieving web pages by reranking search results.
Download
Paper Nr: 230
Title:

TRIGGERING RULES FOR CONVERSATIONAL AGENTS IN TRADING SITUATIONS
Authors:

Grzegorz Dziczkowski, Arnaud Doniec and Stéphane Lecoeuche
Abstract: This paper describes a methodology to establish behavior rules for conversational agents on commercial web sites. Our work is a contribution to a recent research field: agent mining (Cao, 2009) which results from two interrelated research area: Agent/Multi-agent system and Data Mining. The proposed methodology is based on behavior analysis of e-commerce clients and customers’ segmentation. Our proposal has been applied on a real commercial web site to construct the triggering rules of a virtual seller agent.
Download
Paper Nr: 236
Title:

PROPOSAL OF A FRAMEWORK TO SHARE KNOWLEDGE ON CONSUMER’S IMPRESSIONS
Authors:

Keiichi Muramatsu, Tatsuo Togawa, Kazuaki Kojima and Tatsunori Matsui
Abstract: Recently, impressions of web pages formed by visitors have become an important tool to increase the number of repeat visitors to the web page. Therefore, the management of knowledge on consumers’ impressions obtained in several study fields is an essential task in current industrial design. However, no methods that realize such knowledge management have been established. Thus, our study proposes and implements a knowledge management method that can effectively provide knowledge of impressions to web designers to help them in building attractive websites. We introduce a framework for the description of impressions that depend on perceptual fluency, which can serve as an useful indicator of pleasure. We can extract the features of objects that affect impressions on the basis of perceptual fluency. We specify the relationship between objects and impressions by modeling the concepts of awareness, perception, and self-report on the basis of an ontology development environment Hozo and a top-level ontology YAMATO. We then instantiate a case where a person has a good impression of a Web page, and we describe the relationship between a perception and a stimulus in such a case. Our approach demonstrates that ontological modeling of impressions helps us to understand the correspondences between affections and physical irritations.
Download
Paper Nr: 239
Title:

LEARNING FROM DEMONSTRATION - Automatic Generation of Extended Behavior Networks for Autonomous Robots from an Expert’s Demonstration
Authors:

Stefan Czarnetzki, Sören Kerner and Patrick Szcypior
Abstract: The recent research focus on autonomous mobile robots has greatly improved their capability to perform complex tasks, making it more and more difficult to design eligible behavior manually. Therefore this paper presents an algorithm to automatically derive a behavior network from demonstration by an expert. Different tasks are evaluated to proof the generalizability and robustness of the proposed demonstration approach.
Download
Paper Nr: 245
Title:

PREDICTING PERFORMANCE IN TEAM GAMES - The Automatic Coach
Authors:

Guillermo Jiménez-Díaz, Héctor D. Menéndez, David Camacho and Pedro A. González-Calero
Abstract: A wide range of modern videogames involves a number of players collaborating to obtain a common goal. The way the players are teamed up is usually based on a measure of performance that makes players with a similar level of performance play together. We propose a novel technique based on clustering over observed behaviour in the game that seeks to exploit the particular way of playing of every player to find other players with a gameplay such that in combination will constitute a good team, in a similar way to a human coach. This paper describes the preliminary results using these techniques for the characterization of player and team behaviours. Experiments are performed in the domain of Soccerbots.
Download
Paper Nr: 255
Title:

PRICE-SETTING BASED COMBINATORIAL AUCTION APPROACH FOR CARRIERS’ COLLABORATION IN LESS THAN TRUCKLOAD TRANSPORTATION
Authors:

Bo Dai and Haoxun Chen
Abstract: In collaborative logistics, multiple carriers may form an alliance to optimize their transportation operations through sharing transportation requests and vehicle capacities. In this paper, we study a carriers’ collaboration problem in less than truckload transportation with pickup and delivery requests. After formulating the problem as a mixed integer programming model, an iterative price-setting based combinatorial auction approach based on Lagrangian relaxation is proposed. Numerical experiments on randomly generated instances demonstrate the effectiveness of the approach.
Download
Paper Nr: 257
Title:

IN YOUR INTEREST - Objective Interestingness Measures for a Generative Classifier
Authors:

Dominik Fisch, Edgar Kalkowski, Bernhard Sick and Seppo J. Ovaska
Abstract: In a wide-spread definition, data mining is termed to be the “non-trivial process of identifying valid, novel, potentially useful, and ultimately understandable patterns in data”. In real applications, however, usually only the validity of data mining results is assessed numerically. An important reason is that the other properties are highly subjective, i.e., they depend on the specific knowledge and requirements of the user. In this article we define some objective interestingness measures for a specific kind of classifier, a probabilistic classifier based on a mixture model. These measures assess the informativeness, uniqueness, importance, discrimination, comprehensibility, and representativity of rules contained in this classifier to support a user in evaluating data mining results. With some simulation experiments we demonstrate how these measures can be applied.
Download
Paper Nr: 266
Title:

AN APPROACH TO SEMI-SUPERVISED CLASSIFICATION USING THE HUNGARIAN ALGORITHM
Authors:

Amparo Albalate, Aparna Suchindranath and Wolfgang Minker
Abstract: In this paper we propose a novel semi-supervised classification algorithm from the cluster-and-label framework. A small amount of labeled examples is used to automatically label the extracted clusters, so that the initial labeled seed is implicitely ”augmented” to the whole clustered data. The optimum cluster labelling is achieved by means of the Hungarian algorithm, traditionally used to solve any optimisation assignment problem. Finally, the augmented labeled set is applied to train a SVM classifier. This semi-supervised approach has been compared to a fully supervised version. In our experiments we used an artificial dataset (mixture of Gaussians) as well as other five real data sets from the UCI repository. In general, the experimental results showed significant improvements in the classification performance under minimal labeled sets using the semi-supervised algorithm.
Download
Paper Nr: 293
Title:

SOLVING DISTRIBUTED CONSTRAINT OPTIMIZATION PROBLEMS - An Evolutionary Approach
Authors:

Maryam Rahmaninia, Elnaz Bigdeli and Mohsen Afsharchi
Abstract: A significant body of work in multi-agent systems over more than two decades has focused on multi-agent coordination (Levesque et al., 1990). Many challenges in multi-agent coordination can be modeled as Distributed Constraint Optimizations (DCOPs). Many complete and incomplete algorithms have been introduced for DCOPs, but complete algorithms are often impractical for large scale and dynamic environments which lead to the study of incomplete algorithms. In both complete and incomplete algorithms, computational cost is a major concern. Different approaches are introduced to solve this problem and improve existing algorithms. The main contribution of this paper is to decrease computational cost of DALO-t (Distributed Asynchronous Local Optimization) algorithm by introducing a new algorithm to find the best solution. This new algorithm is called Genetic Distributed Asynchronous Local Optimization (GDALO-t). GDALO-t is an effective method to reduce computational load and power consumption in implementation. This paper, under various assumptions, presents an analysis of this new algorithm.
Paper Nr: 295
Title:

CASE REPRESENTATION AND ADAPTATION IN SMARTLP - A Web-based Lesson Planning System
Authors:

Aslina Saad, P. W. H. Chung and C. W. Dawson
Abstract: Lesson plans help teachers to organize content, materials and methods for their teaching. Appropriate lesson plans are crucial to accommodate student differences in various aspects. Currently there are limited mechanisms to support decision making in constructing lesson plans based on the constraints teachers have. Since lesson plans have a standard format, they can potentially be shared. SmartLP, a web-based lesson planning system, was developed to assist teachers in preparing suitable lesson plans based on various constraints; students’ profile, curriculum and facilities. In SmartLP, teachers can make modification to the retrieved plans according to their constraints, as opposed to generating new ones from scratch. Implementation of such systems insists on a proper case representation as it facilitates case retrieval and subsequently case adaptation to handle differences in hand. An ontology for the lesson plan domain has been built in the form of a taxonomy. This is followed by case definition that consists of problem description and solution. Cases are represented as attributes - value representation in a case base. Transformation, a kind of case adaptation, is implemented in the system to facilitate teachers in adding, deleting or editing the contents of the retrieved lesson plans. The adaptation can be derived from one case or several cases.
Download
Paper Nr: 302
Title:

SPEEDING UP LATENT SEMANTIC ANALYSIS - A Streamed Distributed Algorithm for SVD Updates
Authors:

Radim Řehůřek
Abstract: Since its inception 20 years ago, Latent Semantic Analysis (LSA) has become a standard tool for robust, unsupervised inference of semantic structure from text corpora. At the core of LSA is the Singular Value Decomposition algorithm (SVD), a linear algebra routine for matrix factorization. This paper introduces a streamed distributed algorithm for incremental updates, which allows the factorization to be computed rapidly in a single pass over the input matrix on a cluster of autonomous computers.
Download
Paper Nr: 315
Title:

AUTOMATIC CONTENT EXTRACTION ON THE WEB WITH INTELLIGENT ALGORITHMS

Authors:

Pablo Cababie, Alvaro Zweig, Gabriel Barrera and Daniela Lopéz De Luise
Abstract: Since the INTERNET outburst, consumer perception turned into a complex issue to be measured. Non-traditional advertising methods and new product exhibition alternatives emerged. Forums and review sites allow end users to suggest, recommend or rate products according to their experiences. This gave raise to the study of such data collections. After analyze, store and process them properly, they are used to make reports used to assist in middle to high staff decision making. This research aims to implement concepts and approaches of artificial intelligence to this area. The framework proposed here (named GDARIM), is able to be parameterized and handled to other similar problems in different fields. To do that it first performs deep problem analysis to determine the specific domain variables and attributes. Then, it implements specific functionality for the current data collection and available storage. Next, data is analyzed and processed, using Genetic Algorithms to retro feed the keywords initially loaded. Finally, properly reports of the results are displayed to stakeholders.
Paper Nr: 320
Title:

A LINGUISTIC GROUP DECISION MAKING METHOD BASED ON DISTANCES
Authors:

Edurne Falcó and José Luis García-Lapresta
Abstract: It is common knowledge that the political voting systems suffer inconsistencies and paradoxes such that Arrow has shown in his well-known Impossibility Theorem. Recently Balinski and Laraki have introduced a new voting system called Majority Judgement (MJ) which tries to solve some of these limitations. In MJ voters have to asses the candidates through linguistic terms belonging to a common language. From this information, MJ assigns as the collective assessment the lower median of the individual assessments and it considers a sequential tie-breaking method for ranking the candidates. The present paper provides an extension of MJ focused to reduce some of the drawbacks that have been detected in MJ by several authors. The model assigns as the collective assessment a label that minimizes the distance to the individual assessments. In addition, we propose a new tie-breaking method also based on distances.
Download
Paper Nr: 322
Title:

MACHINE SYMBOL GROUNDING AND OPTIMIZATION
Authors:

Oliver Kramer
Abstract: Autonomous systems gather high-dimensional sensorimotor data with their multimodal sensors. Symbol grounding is about whether these systems can, based on this data, construct symbols that serve as a vehicle for higher symbol-oriented cognitive processes. Machine learning and data mining techniques are geared towards finding structures and input-output relations in this data by providing appropriate interface algorithms that translate raw data into symbols. Can autonomous systems learn how to ground symbols in an unsupervised way, only with a feedback on the level of higher objectives? A target-oriented optimization procedure is suggested as a solution to the symbol grounding problem. It is demonstrated that the machine learning perspective introduced in this paper is consistent with the philosophical perspective of constructivism. Interface optimization offers a generic way to ground symbols in machine learning. The optimization perspective is argued to be consistent with von Glasersfeld’s view of the world as a black box. A case study illustrates technical details of the machine symbol grounding approach.
Download
Paper Nr: 324
Title:

INTENTION RECOGNITION WITH EVENT CALCULUS GRAPHS AND WEIGHT OF EVIDENCE
Authors:

Fariba Sadri
Abstract: Intention recognition has significant applications in ambient intelligence, for example in assisted living and care of the elderly, in games and in crime detection. In this paper we describe an intention recognition system based on a formal logic of actions and fluents. The system, called WIREC, exploits plan libraries as well as a basic theory of actions, causality and ramifications. It also exploits profiles, contextual information, heuristics, the actor’s knowledge seeking actions, and any available integrity constraints. Whenever the profile and context suggest there is a usual pattern of behaviour on the part of the actor the search for intention is focused on existing plan libraries. But, when no such information is available or if the behaviour of the actor deviates from the usual pattern, the search for intentions reverts to the basic theory of actions, in effect dynamically constructing possible partial plans corresponding to the actions executed by the actor.
Download
Paper Nr: 376
Title:

DEALING WITH “VERY LARGE” DATASETS - An Overview of a Promising Research Line: Distributed Learning
Authors:

Diego Peteiro-Barral, Bertha Guijarro-Berdiñas and Beatriz Pérez-Sánchez
Abstract: Traditionally, a bottleneck preventing the development of more intelligent systems was the limited amount of data available. However, nowadays in many domains of machine learning, the size of the datasets is so large that the limiting factor is the inability of learning algorithms to use all the data to learn with in a reasonable time. In order to handle this problem a new field in machine learning has emerged: large-scale learning, where learning is limited by computational resources rather than by the availability of data. Moreover, in many real applications, “very large” datasets are naturally distributed and it is necessary to learn locally in each of the workstations in which the data are generated. However, the great majority of well-known learning algorithms do not provide an admissible solution to both problems: learning from “very large” datasets and learning from distributed data. In this context, distributed learning seems to be a promising line of research with which to deal with both situations, since “very large” concentrated datasets can be partitioned among several workstations. This paper provides some background regarding distributed environments as well as an overview of distributed learning for dealing with “very large” datasets.
Download
Paper Nr: 393
Title:

D-RANK: A FRAMEWORK FOR SCORE AGGREGATION IN SPECIALIZED SEARCH
Authors:

Martin Veselý, Martin Rajman, Jean-Yves Le Meur, Ludmila Marian and Jérôme Caffaro
Abstract: In this paper we present an approach to score aggregation for specialized search systems. In our work we focus on document ranking in scientific publication databases. We work with the collection of scientific publications of the CERN Document Server. This paper reports on work in progress and describes rank aggregation framework with score normalization. We present results that we obtained with aggregations based on logistic regression using both ranks and scores. In our experiment we concluded that score-based aggregation favored performance in terms of Average Precision and Mean Reciprocal Rank, while rank-based aggregation favored document discovery.
Download
Paper Nr: 395
Title:

INTERPRETING BULGARIAN SOUND ALTERNATIONS OF INFLECTIONAL MORPHOLOGY IN DATR
Authors:

Velislava Stoykova
Abstract: The paper presents an approach to interpret sound alternations of Bulgarian language for inflectional morphology of definite article. The DATR language for lexical knowledge presentation is accepted as a framework, and the analysis and examples of semantic network for different part-of-speech are presented. Finally, more general conclusions for formal interpretation of sound alternations for inflectional morphology are defined.
Download
Paper Nr: 398
Title:

LogAnswer IN QUESTION ANSWERING FORUMS
Authors:

Björn Pelzer, Ingo Glöckner and Tiansi Dong
Abstract: LogAnswer is a question answering (QA) system for the German language. By providing concise answers to questions of the user, LogAnswer provides more natural access to document collections than conventional search engines do. QA forums provide online venues where human users can ask each other questions and give answers. We describe an ongoing adaptation of LogAnswer to QA forums, aiming at creating a virtual forum user who can respond intelligently and efficiently to human questions. This serves not only as a more accurate evaluation method of our system, but also as a real world use case for automated QA. The basic idea is that the QA system can disburden the human experts from answering routine questions, e.g. questions with known answer in the forum, or questions that can be answered from the Wikipedia. As a result, the users can focus on those questions that really demand human judgement or expertise. In order not to spam users, the QA system needs a good self-assessment of its answer quality. Existing QA techniques, however, are not sufficiently precision-oriented. The need to provide justified answers thus fosters research into logic-oriented QA and novel methods for answer validation.
Download
Paper Nr: 399
Title:

STATISTICAL LANGUAGE IDENTIFICATION OF SHORT TEXTS
Authors:

Fela Winkelmolen and Viviana Mascardi
Abstract: Although correctly identifying the language of short texts should prove useful in a large number of applications, few satisfactory attemps are reported in the literature. In this paper we describe a Naive Bayes Classifier that performs well on very short texts, as well as the corpus that we created from movie subtitles for training it. Both the corpus and the algorithm are available under the GNU Lesser General Public License.
Download
Paper Nr: 402
Title:

LOCATING INFORMATION-HIDING IN MP3 AUDIO
Authors:

Mengyu Qiao, Andrew H. Sung, Qingzhong Liu and Bernardete M. Ribeiro
Abstract: Steganography provides a stealthy communication channel for malicious users, which jeopardizes traditional cyber security infrastructure. Due to the good quality and the small storage usage, compressed audio has been widely employed by online audio sharing, audio streaming broadcast, and voice over IP, etc. Several audio steganographic systems have been developed and published on Internet. Traditional blind steganalysis methods detect the existence of information hiding, but neglect the size and the location of hidden data. In this paper, we present a scheme to locate the modified segments in compressed audio streams based on signal analysis in MDCT transform domain. We create reference signals by reversing and repeating quantification process, and compare the statistical differences between source signals and reference signals. Dynamic evolving neural-fuzzy inference systems are applied to predict the number of modified frames. Finally, the frames of audio streams are ranked according to their modification density, and the top ranked frames are selected as candidate information-hiding locations.
Download
Paper Nr: 410
Title:

MECHANISMS FOR TEMPORAL LOGIC IMPLEMENTATION IN RULE-BASED SYSTEMS
Authors:

Josef Hahn, Karl-Heinz Krempels and Christoph Terwelp
Abstract: Rule-based systems are more and more important in middleware architectures and distributed applications. Although support for temporal constructs would be very conveniant for many domains, implementations are not yet widespread. This paper is about several methods to expand rule-based systems and the commonly used RETE algorithm in order to gain basic support for temporal logic constructs. A few promising approaches are discussed and compared with respect to efficiency, memory usage, and implemetation details. The paper is limited on the discussion of temporal logics in rule-based systems and does not take temporal logic in other contexts into account.
Download
Paper Nr: 411
Title:

PETRI NET BASED AGENTS FOR COORDINATING RESOURCES IN A WORKFLOW MANAGEMENT SYSTEM
Authors:

Albert Plà, Pablo Gay, Joaquim Meléndez and Beatriz López
Abstract: We present a new framework for business process management based in a Petri net extension called Resource-Aware Petri Nets. This extension considers resources representation at the application level and allows the monitoring of the whole system with its dependencies. Thus, to solve resource usage conflict, agents are proposed to take care of monitoring workflow instances. This new comprehension of dependencies also allows the creation of a delay prediction system based in historical data from the workflows itself. In this paper we expose our methodology for modeling workflows through our extension which is based in classical approaches. Also a monitoring and delay prediction workflow is introduced and analyzed. In order to test our approach, we have extracted workflows from real cases and tested our framework simulating different kind of situations and resources, getting promising results since our prototype can provide early detection of workflows delays.
Download
Paper Nr: 412
Title:

MULTI-AGENT BASED SURVEILLANCE OF HUMAN WORKFLOWS
Authors:

Emmanuel Sardis, Vasileios Anagnostopoulos and Theodora Varvarigou
Abstract: Workflow recognition through processing of humans and objects in a camera sensor network, presents a significant challenge recently. Human action recognition and sequence of actions manipulation, that construct a workflow situation/rule, has many practical applications in many different real human application environments. This article presents a multi agent based real time infrastructure, for recognizing humans workflows, by evaluating and processing computer vision signals, from multiple cameras sensors. The system architecture, the related agents’ infrastructure of the distributed environment of sensors, are presented together with the algorithmic modules, that evaluate sensors signals into workflow events, and related alarms’ outputs from the system. The article presents a full functional system that integrates the distributed functionality of the multi agents’ infrastructure into real working environments, using the JADE agents’ technologies. The evaluations of system simulation results are conclude this work, giving related feedback for possible future architecture and implementation extensions.
Download
Paper Nr: 413
Title:

EXPLOITING VISUAL OBSERVATIONS FOR EFFICIENT WORKFLOW SCHEDULING IN PRODUCTION ENVIRONMENTS
Authors:

Anastasios Doulamis
Abstract: This paper proposes a new production scheduling algorithm that exploits (a) visual observations of industrial operations to estimate the actual completion times for tasks and (b) incremental graph partitioning-based clustering algorithms. The latter are implemented through an incremental implementation of the spectral clustering. Computer vision tools are applied to identify industrial operations via visual observations.
Download
Paper Nr: 56
Title:

A CONTINUOS LEARNING FOR A FACE RECOGNITION SYSTEM
Authors:

Aldo F. Dragoni, Germano Vallesi and Paola Baldassarri
Abstract: A system of Multiple Neural Networks has been proposed to solve the face recognition problem. Our idea is that a set of expert networks specialized to recognize specific parts of face are better than a single network. This is because a single network could no longer be able to correctly recognize the subject when some characteristics partially change. For this purpose we assume that each network has a reliability factor defined as the probability that the network is giving the desired output. In case of conflicts between the outputs of the networks the reliability factor can be dynamically re-evaluated on the base of the Bayes Rule. The new reliabilities will be used to establish who is the subject. Moreover the network disagreed with the group and specialized to recognize the changed characteristic of the subject will be retrained and then forced to correctly recognize the subject. Then the system is subjected to continuous learning.
Download
Paper Nr: 76
Title:

MODELLING A BACKGROUND FOR BACKGROUND SUBTRACTION FROM A SEQUENCE OF IMAGES - Formulation of Probability Distribution of Pixel Positions
Authors:

Suil Son, Young-Woon Cha and Suk I. Yoo
Abstract: This paper presents a new background subtraction approach to identifying the various changes of objects in a sequence of images. A background is modelled as the probability distribution of pixel positions given intensity clusters, which is constructed from a given sequence of images. Each pixel position in a new image is then identified with either a background or a foreground, depending on its value from probability distribution of pixel positions representing a background. The presented approach is illustrated using two examples. As compared to traditional intensity-based approaches, this approach is shown to be robust to dynamic textures and various changes of illumination.
Download
Paper Nr: 77
Title:

CONVERSATIONAL AGENT IN ARGUMENTATION - A Model and Evaluation on a Dialogue Corpus
Authors:

Mare Koit
Abstract: Communication between two participants, A and B, is considered, where A has a communicative goal that his/her partner, B, will make a decision to perform an action D. A computational model of argumentation is developed which includes reasoning. Communicative strategies and tactics used by participants for achieving their communicative goals are considered. A simple dialogue system (conversational agent) is implemented which can optionally play the role of A or B using classified sets of pre-defined Estonian sentences. For further evaluation of the model and with the aim to develop the dialogue system, the analysis of the Estonian Dialogue Corpus is carried out. Calls of sales clerks who persuade clients to take training courses of an educational company are analysed. The calls end mostly with the postponement of the decision therefore the sales clerks do not achieve their communicative goal.
Download
Paper Nr: 102
Title:

DATA ANALYSIS OF AGE-RELATED CHANGES IN VISUAL MOTION PERCEPTION
Authors:

Nadejda Bocheva, Olga Georgieva and Miroslava Stefanova
Abstract: Many cognitive abilities decline with age, but ageing is accompanied by great variability within older population. The aim of the present study is to explore the possibility to differentiate the age-related and the individual differences in visual information processing. Two different analytical methods – mixed ANOVA and fuzzy clustering, were applied to the data of psychophysical experiments on motion direction discrimination. The results suggest that the complementary analysis based on both methods offers new opportunities to retrieve information from the psychophysical studies and to separate the differences due to age and gender from the individual differences of the participants. The proposed data analytic approach allows better understanding of the factors that caused variation in performance with age and can be used as a diagnostic tool to distinguish pathological from normal ageing.
Download
Paper Nr: 108
Title:

A PLANNING AND SCHEDULING PERSPECTIVE FOR DESIGNING BUSINESS PROCESSES FROM DECLARATIVE SPECIFICATIONS
Authors:

Irene Barba and Carmelo Del Valle
Abstract: Usually, business process models are manually achieved by business analysts and most of current modelling languages are of imperative nature. As a consequence, non-optimized or faulty models can be obtained. This work proposes a planning based approach to give business analysts assistance for the process models generation. This approach entails the selection and the order of the activities to be executed (planning), and the resources allocation involving temporal reasoning (scheduling), both considering function optimization. The process information is specified in a declarative way, that is translated into the standard planning language PDDL. A friendly graphic language is used (ConDec-R, an extension of ConDec).
Download
Paper Nr: 117
Title:

TAG RECOMMENDATION BASED ON USER’S BEHAVIOR IN COLLABORATIVE TAGGING SYSTEMS
Authors:

Nagehan Ilhan and Şule Gündüz-Öğüdücü
Abstract: Social bookmarking Web sites allow users submitting their resources and labeling them with arbitrary keywords, called tags, to create folksonomies. Tag recommendation is an important element of collaborative tagging systems which aims at providing relevant information to users by proposing a set of tags to each newly posted resource. In this paper, we focus on the task of tag recommendation when a user examines a document based on the user’s tagging behavior. We explore the use of this semantic relationship in modeling the user tagging behavior. The experiments are performed on the data set obtained from a social bookmarking site. Our experimental result show that our method is efficient in modeling users’ tagging behavior and it can be used to recommend tags for resources.
Download
Paper Nr: 123
Title:

MINING QUANTITATIVE ASSOCIATION RULES IN MICROARRAY DATA USING EVOLUTIVE ALGORITHMS
Authors:

Maria Martinez Ballesteros, Cristina Rubio Escudero, J. C. Riquelme and F. Martíınez-Álvarez
Abstract: The microarray technique is able to monitor the change in concentration of RNA in thousands of genes simultaneously. The interest in this technique has grown exponentially in recent years and the difficulties in analyzing data from such experiments, which are characterized by the high number of genes to be analyzed in relation to the low number of experiments or samples available. In this paper we show the result of applying a data mining method based on quantitative association rules for microarray data. These rules work with intervals on the attributes, without discretizing the data before. The rules are generated by an evolutionary algorithm.
Download
Paper Nr: 137
Title:

CONSISTENCY AT THE CORE OF COMMONSENSE
Authors:

Donald Perlis
Abstract: This paper argues for a "commonsense core" hypothesis, with emphasis on the issue of consistency in agent knowledge bases. This is part of a long-term research program, in which the hypothesis itself is being gradually refined, in light of various sorts of evidence. The gist is that a commonsense reasoning agent that would otherwise become incapacitated in the presence of inconsistent data may – by means of a modest additional error-handling “core” component – carry out more effective real-time reasoning, and also that there may be cases of interest in which the “core” is more usefully integrated into the knowledge base itself.
Download
Paper Nr: 159
Title:

A LOCAL SEARCH APPROACH TO SOLVE INCOMPLETE FUZZY CSPs
Authors:

Mirco Gelain, Maria Silvia Pini, Francesca Rossi, Kristen Brent Venable and Toby Walsh
Abstract: We consider fuzzy constraint problems where some of the preferences may be unspecified. This models, for example, settings where agents are distributed and have privacy issues, or where there is an ongoing preference elicitation process. In this context, we study how to find an optimal solution without having to wait for all the preferences. In particular, we define local search algorithms that interleave search and preference elicitation, with the goal to find a solution which is ”necessarily optimal”, that is, optimal no matter what the missing data are, while asking the user to reveal as few preferences as possible. While in the past this problem has been tackled with a branch & bound approach, which was guaranteed to find a solution with this property, we now want to see whether a local search approach can solve such problems optimally, or obtain a good quality solution, with fewer resources. At each step, our local search algorithm moves from the current solution to a new one, which differs in the value of a variable. The variable to reassign and its new value are chosen so to maximize the quality of the next solution. To compute this, we elicit some of the missing preferences in the neighbor solutions. Experimental results on randomly generated fuzzy CSPs with missing preferences show that our local search approach is promising, both in terms of percentage of elicited preferences and scaling properties.
Download
Paper Nr: 181
Title:

SINGLE-WALK PARALLELIZATION OF THE GENETIC ALGORITHM
Authors:

Wojciech Bożejko and Mieczyslaw Wodecki
Abstract: This paper aims at presenting theoretical properties which can be used to approximate the theoretical speedup of parallel genetic algorithms. The most frequently parallelization method employed to genetic algorithm implements a master-slave model by distributing the most computationally exhausting elements of the algorithm (usually evaluation of the fitness function, i.e. cost function calculation) among a number of processors (slaves). This master-slave parallelization is regarded as easy in programming, which makes it popular with practitioners. Additionally, if the master processor keeps the population (and slave processors are used only as computational units for individuals fitness function evaluation), it explores the solution space in exactly the same manner as the sequential genetic algorithm. In this case we can say that we analyze the single-walk parallel genetic algorithm.
Download
Paper Nr: 189
Title:

OPPONENT-BASED TACTIC SELECTION FOR A FIRST PERSON SHOOTER GAME

Authors:

David Thomson
Abstract: Video games are quickly becoming a significant part of society with a growing industry that employs a wide range of talent, from programmers to graphic artists. Video games are also becoming an interesting and useful testbed for Artificial Intelligence research. Complex, realistic environmental constraints, as well as performance considerations demand highly efficient AI techniques. At the same time, the AI component of a video game may define the ongoing commercial success, or failure, of a particular game or game engine. This research details an approach to opponent modeling in a first person shooter game, and evaluates proficiency gains facilitated by such a technique. Information about the user is recorded and used by the existing Artificial Intelligence component to select tactics for any given opponent. The evaluation results show that when computer characters use such modeling they are more effective than when they do not model their opponent.
Download
Paper Nr: 203
Title:

SURFACE ROUGHNESS MODELLING AND OPTIMIZATION IN CNC END MILLING USING TAGUCHI DESIGN AND NEURAL NETWORKS

Authors:

Menelaos Pappas, John Kechagias, Vassilis Iakovakis and Stergios Maropoulos
Abstract: A Neural Network modelling approach is presented for the prediction of surface texture parameters during end milling of aluminium alloy 5083. Eighteen carbide end mill cutters were manufactured by a five axis grinding machine and assigned to mill eighteen pockets having different combinations of geometry parameters and cutting parameter values, according to the L18 (21x37) standard orthogonal array. A feed-forward back-propagation NN was developed using data obtained from experimental work conducted on a CNC milling machine center according to the principles of Taguchi’s design of experiments method. It was found that NN approach can be applied easily on designed experiments and predictions can be achieved, fast and quite accurately.
Download
Paper Nr: 215
Title:

AN INTEGRATED SYSTEM FOR SCHOOL TIMETABLING
Authors:

Luisa Carpente, Ana Cerdeira-Pena, Guillermo de Bernardo and Diego Seco
Abstract: In this paper, we present an application that covers the whole complex school timetabling process, from the data introduction to the final adjustment of the automatically generated solution. On the one hand, our application interacts with the Academic Administration Official Systems (AAOS) and simplifies the hard phase of introducing the data. On the other hand, complete solutions are efficiently provided by an algorithmic engine based on different heuristic techniques, and easily updated by means of a thoroughly designed user interface.
Download
Paper Nr: 218
Title:

AUTOMATIC FACE RECOGNITION - Methods Improvement and Evaluation
Authors:

Ladislav Lenc and Pavel Král
Abstract: This paper deals with Automatic Face Recognition (AFR), which means automatic identification of a person from a digital image. Our work focuses on an application for Czech News Agency that will facilitate to identify a person in a large database of photographs. The main goal of this paper is to propose some modifications and improvements of existing face recognition approaches and to evaluate their results. We assume that about ten labelled images of every person are available. Three approaches are proposed: the first one, Average Eigenfaces, is a modified Eigenfaces method; the second one, SOM with Gaussian mixture model, uses Self Organizing Maps (SOMs) for image reduction in the parametrization step and a Gaussian Mixture Model (GMM) for classification; and in the last one, Re-sampling with a Gaussian mixture model, several resize filters are used for image parametrization and a GMM is also used for classification. All experiments are realized using the ORL database. The recognition rate of the best proposed approach, SOM with Gaussian mixture model, is about 97%, which outperforms the “classic” Eigenfaces, our baseline, by 27% in absolute value.
Download
Paper Nr: 220
Title:

A GAME PLAYING ROBOT THAT CAN LEARN A TACTICAL KNOWLEDGE THROUGH INTERACTING WITH A HUMAN
Authors:

Raafat Mahmoud, Atsushi Ueno and Shoji Tatsumi
Abstract: We propose a new approach for teaching a humanoid-robot a task online without pre-set data provided in advance. In our approach, human acts as a collaborator and also as a teacher. The proposed approach enables the humanoid-robot to learn a task through multi-component interactive architecture. The components are designed with the respect to human methodology for learning a task through empirical interactions. For efficient performance, the components are isolated within one single API. Our approach can be divided into five main roles: perception, representation, state/knowledge-up-dating, decision making and expression. A conducted empirical experiment for the proposed approach is to be done by teaching a Fujitsu’s humanoid-robot "Hoap-3" an X-O game strategy and its results are to be done and explained. Important component such as observation, structured interview, knowledge integration and decision making are described for teaching the robot the game strategy while conducting the experiment.
Download
Paper Nr: 226
Title:

A “CONTEXT EVALUATOR” MODEL FOR A MULTIMODAL USER INTERFACE IN DRIVING ACTIVITY
Authors:

Jesús Murgoitio, Arkaitz Urquiza, Maider Larburu and Javier Sánchez
Abstract: Focused on new HMI for vehicles, this paper explains the methodology followed by Tecnalia to provide an evaluation of the context, based on sensors data, in order to achieve the optimum usability and ubiquity levels at any given moment, and finally to optimize the driving task from the safety point of view. The paper is divided into three sections. Firstly, there is an introduction in which the main concepts, multimodality and Ambient intelligence, are explained (section 1), and this is followed by the two projects related to the work shown here (section 2). In section 3 we will go on to describe the methodology used to obtain an evaluation of the context, as is necessary for a car HMI with multimodal capabilities. Finally, a particular “Context evaluator” is displayed in order to illustrate the methodology and report the work carried out.
Download
Paper Nr: 232
Title:

TRENDSPOTTER DETECTION SYSTEM FOR TWITTER
Authors:

Wataru Shirakihara, Tetsuya Oishi, Ryuzo Hasegawa, Hiroshi Hujita and Miyuki Koshimura
Abstract: It is too difficult for us to find out trends with search engines. Twitter, a popular microblogging tool, has seen a lot of growth since it launched in October, 2006. Information about the trends is posted by many twitterers. If we find out trendspotters from twitterers, and follow them, we can get it more easily. Our trendspotter detection system uses the burst detection algorithm, and we verified its effectiveness for Twitter’s posts. We succeeded in detecting the 24 trendspotters by 5277 users.
Download
Paper Nr: 235
Title:

SMOOTHED HEX-GRID TRAJECTORY PLANNING USING HELICOPTER DYNAMICS
Authors:

Lukáš Chrpa and Antonín Komenda
Abstract: Considering Unmanned Autonomous Vehicles (UAVs) the planning tasks mainly consist of finding paths between given waypoints with respect to given constraints. In this paper we developed a path planning system for flying UAVs (VTOLs and CTOLs) built upon Hexagonal grids which also supports simple dynamics (handling with speed). The planning system is additionally supported by a trajectory smoothing mechanism based on a dynamics model of a helicopter. The model can be also used for the simulation of the helicopter movement.
Download
Paper Nr: 243
Title:

SMART DOCUMENT TECHNOLOGIES FOR EXTRACTING AND STRUCTURING DATA FROM PATIENT RECORDS - Opportunities for New Knowledge based Services
Authors:

Denys Proux, Eric Cheminot, Caroline Hagege and Frederique Segond
Abstract: Costs related to healthcare are exploding. In Europe studies show that they will reach in 2050 up to 10% to 15% of the Gross Domestic Product. Several causes drive up these costs such as ageing population or the rise of chronic diseases. Adverse events jeopardizing patient safety in complex hospital workflows such as Hospital Acquired Infections are also a major contributor to these costs. But recent studies suggest that it might be possible to reduce this impact by 20 to 30% with a combination of appropriate processes and smart monitoring tools. Information Technologies are therefore seen as a key enabler to help improving information workflow and process efficiency inside hospitals.
Download
Paper Nr: 248
Title:

REAL-TIME CONTEXT AWARE REASONING IN ON-BOARD INTELLIGENT TRAFFIC SYSTEMS - An Architecture for Ontology-based Reasoning using Finite State Machines
Authors:

Arjan Stoter, Simon Dalmolen, Eduard Drenth, Eric Cornelisse and Wico Mulder
Abstract: In-vehicle information management is vital in intelligent traffic systems. In this paper we motivate an architecture for ontology-based context-aware reasoning for in-vehicle information management. An ontology is essential for system standardization and communication, and ontology-based reasoning allows context-awareness, inference and advanced reasoning capabilities. However, the amount of computational power it requires often conflicts with the computational limitations of on-board units, as well as the high demand for timeliness and safety. Our approach uses ontology-based reasoning and a finite state machine (FSM). By combining ontology and FSM, we illustrate how a heavy-weight reasoning-solution could be applied in a light-weight computational environment.
Download
Paper Nr: 253
Title:

THE METHODOLOGY OF PARALLEL MEMETIC ALGORITHMS DESIGNING
Authors:

Wojciech Bożejko and Mieczyslaw Wodecki
Abstract: The paper presents the methodology of parallel algorithm designing based on the memetic approach (Lamarck and Baldwin evolution theory) making use of specific properties of the problem and distributed island model. This approach is presented on the example of the single machine scheduling problem with earliness/tardiness penalties.
Download
Paper Nr: 258
Title:

DATA MINING ON THE INSTALLED BASE INFORMATION - Possibilities and Implementations
Authors:

Rashid Bakirov and Christian Stich
Abstract: Managing the installed base at customer sites is a key for customer satisfaction. Hereby installed base comprises installed systems and products at customer sites which are currently being serviced by the producer company. The purpose of the present study is developing use cases for data mining on the installed base information of a large manufacturing company and specifically ABB, and constructing data mining models for their implementation. The aim is to use the available information to enhance customer-tailored sales and proactive service. This includes recommendations to customers and failure prediction. The developed models employ association rules mining, classification and regression, realized with the help of data mining tools Oracle Data Mining and Weka. Results have been evaluated using statistical means, as well as discussed with the experts at the company. These results suggest that with the reasonable amount of data, installed base information is a potential source for data mining models useful for business intelligence.
Download
Paper Nr: 259
Title:

ARTIFACT AS SPECIES - A Formal Approach of the Evolutionary Design
Authors:

Francesco Mele, Antonio Sorgente and Giuseppe Vettigli
Abstract: This paper is focused on everyday artifacts (glasses, pens, pots, chairs etc), that is, objects that have functionality and behaviours which can be associated to the form of individual parts and to the relations between the parts themselves. We formalized the notion of artifact as species, where each species is represented by a class of a formal ontology and individuals by instances of such a class. An evolutionary stage is represented by a relation species-subspecies (corresponding to a class-subclass relation of a formal ontology). To each species is associated a population of individuals (represented by instances of classes). In our approach evolution is represented by a revision of species, hence by a revision of classes, and therefore, by an evolution of an ontology.
Paper Nr: 262
Title:

A HIGH-SPEED ARCHITECTURE FOR BUILDING HYBRID MINDS
Authors:

Oisín Mac Fhearaí, Mark Humphrys and Ray Walshe
Abstract: A resurgence of interest has taken place supporting the idea of an intelligence composed of many simple components or "subminds". There is a growing consensus that, rather than a small number of "elegant" techniques for reasoning, inference or learning being the key to A.I., a model more likely to succeed would consist of perhaps thousands of simple agents co-operating such that an emergent intelligence is seen. The World-Wide Mind project is our attempt to facilitate this approach to scaling up artificial intelligence by enabling large hybrid systems to be built by multiple authors. The goal of the research described in this paper is to improve greatly the speed of the World-Wide Mind platform with a new communications protocol and implementation, to improve the user API and human interfaces, and to investigate methods of automatically constructing effective hybrid minds from the work of multiple authors, as well as to encourage collaboration between A.I. researchers worldwide.
Download
Paper Nr: 277
Title:

INTERVAL BASED INTEGRATED REAL-TIME COORDINATION FOR MULTI-AGENT SYSTEMS
Authors:

Ghulam Mahdi and Abdelkader Gouaich
Abstract: Real-time computations in agent based simulations and (serious) games possess an inherent element of temporal relationships as well as time constraints for their performance and utility measures. Such time relationships and temporal constraints can be observed in individual agent behaviors as well as coordination process involving multiple agent. The temporal relationships and time constraints in multi-agent coordination come in terms of message passing, resource management and negotiations. The idea behind such temporal relationships and time constraints is to efficiently handle complex interactions as different patterns of coordination emerge as per the updated situations under certain time durations. Here we propose our position about integrating both dimensions of individual and collective coordination in a unified manner where the coordination patterns are expressed through Allen’s interval paradigm. We also introduce concept of “timers” to ensure real-time with explicit expressiveness of the interval paradigm.
Download
Paper Nr: 280
Title:

SEMANTIC OBJECT RECOGNITION USING CLUSTERING AND DECISION TREES
Authors:

Falk Schmidsberger and Frieder Stolzenburg
Abstract: Each object in a digital image is composed of many patches (segments) with different shapes and colors. In order to recognize an object, e.g. a table or a book, it is necessary to find out which segments are typical for which object and in which segment neighborhood they occur. If a typical segment in a characteristic neighborhood is found, this segment will be part of the object to be recognized. Typical adjacent segments for a certain object define the whole object in the image. Following this idea, we introduce a procedure that learns typical segment configurations for a given object class by training with example images of the desired object, which can be found in and downloaded from the Internet. The procedure employs methods from machine learning, namely k-means clustering and decision trees, and from computer vision, e.g. contour signatures.
Download
Paper Nr: 312
Title:

DO ARTIFICIAL GENERAL INTELLIGENT SYSTEMS REALLY NEED TO BE CONSCIOUS?
Authors:

J. Ignacio Serrano and M. Dolores del Castillo
Abstract: Consciousness has been studied for long time from heterogeneous perspectives and knowledge fields. In spite of the great numbers of debates and the huge amount of work, the matter is still full of questions and even enigmas at different levels. In addition to well known issues such as the evolutionary utility of consciousness, the existence (or not) of the “hard problem” and the definition itself (to mention just a few), consciousness has also been brought into the mind/body problem. This controversy has encouraged many researchers to tackle the simulation and implementation of consciousness, thus giving rise to the so-called Machine Consciousness (also Artificial Consciousness), which in turn motivates the inclusion of consciousness in Artificial General Intelligent (AGI) Systems. However, do an AGI system need consciousness in order to be (general) intelligent? This paper poses a humble reflection on this subject with the only aim of making the readers think about it before starting working.
Download
Paper Nr: 314
Title:

CREATING CHARACTER CONNECTIONS FROM MANGA
Authors:

Harumi Murakami, Ryota Kyogoku and Hiroshi Ueda
Abstract: We presented a method to create character connections from manga using the frequencies of characters and their co-occurrences by referring to frames. First, we input characters and frames with a data input tool. Second, we calculated the frequencies of characters and the relationships among characters and group-related characters. Third, we created character connections. Preliminary experiments using Dragon Ball vol. 32 suggest the usefulness of our approach.
Download
Paper Nr: 316
Title:

LEUKOCYTES CLASSIFICATION USING BAYESIAN NETWORKS
Authors:

Verónica Rodríguez-López and Raúl Cruz-Barbosa
Abstract: In this paper, the use of bayesian networks in the leukocytes classification problem is explored. The complexity in this problem is mainly due to morphological diversity between cells of the same type and similar features found in different types of cells, which complicate the classification task. Since bayesian networks have demonstrated to be useful as both a classifier and a powerful tool for knowledge representation and inference under conditions of uncertainty, this graphical model is applied in the leukocytes classification problem. The design of two bayesian network models based on the expert’s knowledge and data are presented. Some preliminary results have shown that the proposed models classify all types of leukocytes with an acceptable accuracy.
Download
Paper Nr: 373
Title:

DECISION MAKING BASED ON DUALITY BETWEEN POSITIVE AND NEGATIVE EVALUATIONS
Authors:

Rumiko Azuma, Hayao Miyagi and Yui Miyagi
Abstract: This paper proposes a model of dual hierarchy to study a decision-making structure with positive and negative. It is necessary to evaluate a decision in negative stand point, such as insufficient, hatred, pressure, as a cause of fraud, as well as positive evaluation. In order to treat the positive and negative elements, we propose a model of dual hierarchy process which can evaluate from positive and negative points. Moreover, a technique to judge a consistency of dual evaluation is presented, using a concept of reachability matrix.
Download
Paper Nr: 375
Title:

A PERSONALIZED RECOMMENDER SYSTEM FOR TELECOM PRODUCTS AND SERVICES

Authors:

Zui Zhang, Kun Liu, William Wang, Tai Zhang and Jie Lu
Abstract: The Internet brings excellent opportunities to businesses for providing personalized online services to their customers. Recommender systems are designed to automatically generate personalized recommendations of products and services. This study develops a hybrid recommendation approach which combines user-based and item-based collaborative filtering techniques for mobile product and service recommendation. It particularly implements the approach into an intelligent recommendation system called telecom product recommender system (TCPRS). Experimental results show that the TCPRS can effectively help new customer selecting the most suitable mobile products and services.
Download
Paper Nr: 377
Title:

CONTINGENT PLANNING AS BELIEF SPACE SEARCH
Authors:

Incheol Kim and Hyunsik Kim
Abstract: In this paper, we present a new heuristic search algorithm for solving contingent planning problems with the partial initial condition and nondeterministic actions. The algorithm efficiently searches through a cyclic AND-OR graph with dynamic updates of heuristic values, and generates a contingent plan that is guaranteed to achieve the goal despite of the uncertainty in the initial state and the uncertain effects of actions. Through several experiments, we demonstrate the efficiency of this algorithm.
Download
Paper Nr: 385
Title:

TURNING ARTIFICIAL NEURAL NETWORKS INTO A MARKETING SCIENCE TOOL - Modelling and Forecasting the Impact of Sales Promotions
Authors:

Ibrahim Zafar Qureshi, Marwan Khammash and Konstantinos Nikolopoulos
Abstract: In this study we model the effect of promotions in time-series data and we consequently forecast that extraordinary effect via Artificial Neural Networks (ANN) as implemented from the Expert Method of a popular Artificial Intelligence software. We simulate data considering five factors as to determine the actual impact of each individual promotion. We consider additive and multiplicative models, with the later presenting both linear and non-linear relationships between those five factors; in addition, we superimpose either low or high levels of noise. Our empirical findings suggest that, for nonlinear models with high level of noise, ANN outperform all benchmarks. Standard ANN topologies work well for models with up to two factors while the Expert method provided by the software works well for higher number of factors.
Download
Paper Nr: 390
Title:

DISPERSION EFFECT ON GENERALISATION ERROR IN CLASSIFICATION - Experimental Proof and Practical Algorithm
Authors:

Benoît Gandar, Gaëlle Loosli and Guillaume Deffuant
Abstract: Recent theoretical work proposes criteria of dispersion to generate learning points. The aim of this paper is to convince the reader, with experimental proofs, that dispersion is a good criterion in practice for generating learning points for classification problems. Problem of generating learning points consists then in generating points with the lowest dispersion. As a consequence, we present low dispersion algorithms existing in the literature, analyze them and propose a new algorithm.
Download
Paper Nr: 391
Title:

DECISION SUPPORT TO POLYMER MATERIAL SELECTION
Authors:

Urška Sancin and Bojan Dolšak
Abstract: To succeed means to develop and produce a product with optimal properties for considerable lower price in comparison to similar products on the market. Material selection is one of crucial decisions in product development process affecting quality as well as price of future product. In technical praxis, the designer has to evaluate the information gathered from material data sheets and simulations, engineering analysis and animations of future product performance in virtual environment. Afterwards, he or she has to seek interdependences between them and finally choose the optimum from the broad list of materials. Wide spectrum of various polymers at disposal should be outlined here, as it presents a problem to the designer at polymer material selection process. The proposed decision support system model is an attempt to solve this dilemma and will focus on function, technical features and shape of developing product. Other criteria, like serviceability, technical feasibility and economic justification are going to be considered accordingly. The major benefits concern inexperienced designers along with small and medium sized enterprises (SMEs’).
Download
Paper Nr: 392
Title:

REFLECTIONS ON NEUROCOMPUTATIONAL RELIABILISM
Authors:

Marcello Guarini, Joshua Chauvin and Julie Gorman
Abstract: Reliabilism is a philosophical theory of knowledge that has traditionally focused on propositional knowledge. Paul Churchland has advocated for a reconceptualization of reliabilism to “liberate it” from propositional attitudes (such as accepting that p, believing that p, knowing that p, and the like). In the process, he (a) outlines an alternative for the notion of truth (which he calls “representational success”), (b) offers a non-standard account of theory, and (c) invokes the preceding ideas to provide an account of representation and knowledge that emphasizes our skill or capacity for navigating the world. Crucially, he defines reliabilism (and knowledge) in terms of representational success. This paper discusses these ideas and raises some concerns. Since Churchland takes a neurocomputational approach, we discuss our training of neural networks to classify images of faces. We use this work to suggest that the kind of reliability at work in some knowledge claims is not usefully understood in terms of the aforementioned notion of representational success.
Download
Paper Nr: 400
Title:

DEALING WITH COALITION FORMATION IN THE ROBOCUP RESCUE - An Heuristic Approach
Authors:

Daniel Epstein and Ana L. C. Bazzan
Abstract: Finding an optimal coalition structure to divide agents in groups is equivalent to the set-partitioning problem. Several algorithms have been proposed. However, even to find a sub-optimal value they have to search within an exponential number of coalition structures. Therefore, in this paper, we use one of the proposed algorithms, which is anytime and hence suits environments such as the RoboCup Rescue, where a response is needed in a short time frame. Moreover, we propose to combine this algorithm with heuristics to reduce and constraint the number of agents and tasks that are allowed to participate in a coalition. In this paper we discuss the application of such an approach in a complex task allocation scenario, the RoboCup Rescue.
Download
Paper Nr: 403
Title:

REACTIVE LAYER IN AGI AGENT - Implementation of Adaptive Reactive Behavior and Beyond
Authors:

Vilem Benes
Abstract: Basic mechanisms of cognition working in AGI agent are presented. I argue that reactive behavior is the baseline of intelligence – it is the base component working and it can be further extended to produce more intelligent agents. Mechanisms employed at reactive level enable the agent to develop behavior which both explores and exploits the environment with the purpose of receiving highest reward possible. Three funda-mental mechanisms are intertwined – action selection, action value estimation and situation discrimination. Whole process of adaptation is completely unsupervised and depends only on reward received from envi-ronment. Some technical details of implementation of given mechanisms (BAGIB agent) are described to-gether with implications to other planned parts of “AGI-compliant” architecture. Discussed are several chal-lenges we encounter in AGI, which are not present in usually narrow and domain-limited approach to AI.
Download

Area 2 - Agents

Full Papers
Paper Nr: 28
Title:

AN ABM OF THE DEVELOPMENT OF SHARED MEANING IN A SOCIAL GROUP

Authors:

Enrique Canessa, Sergio E. Chaigneau and Ariel Quezada
Abstract: Generally, concepts are treated as individual-level phenomena. Here, we develop an ABM that treats concepts as group-level phenomena. We make simple assumptions: (1) Different versions exist of one similar conceptualization; (2) When we infer that our view agrees with someone else’s view, we are subject to true agreement (i.e., we really share the concept), but also to illusory agreement (i.e., we do not really share the concept); (3) Regardless whether agreement is true or illusory, it strengthens a concept’s salience in individual minds, and increases the probability of seeking future interactions with that person or source of information. When agents interact using these rules, our ABM shows that three conditions exist: (a) All versions of the same conceptualization strengthen their salience; (b) Some versions strengthen while others weaken their salience; (c) All versions weaken their salience. The same results are corroborated by developing probability models (conditional and Markov chain). Sensitivity analyses to various parameters, allow the derivation of intuitively correct predictions that support our model’s face validity. We believe the ABM and related mathematical models may explain the spread or demise of conceptualizations in social groups, and the emergence of polarized social views, all important issues to sociology and psychology.
Download
Paper Nr: 67
Title:

A COALITION BASED INCENTIVE MECHANISM FOR P2P CONTENT DISTRIBUTION SYSTEMS
Authors:

M. V. Belmonte, M. Díaz and A. Reyna
Abstract: P2P systems suffer from free-loaders, peers that consume many more resources or contents (bandwidth) than they contribute. One of the reasons for this is that the mechanisms used for downloading and sharing in the P2P systems, do no take selfish behavior of the peers into account at the design stage. Therefore, it is important to find mechanisms that provide incentives and encourage cooperative behavior among the peers. One possible solution could be to use an economic framework that provides them with incentives. We propose the application of a coalition formation scheme based on game theory to P2P file sharing systems. The main idea for the coalition formation scheme is based on the fact that peers that contribute more get a better quality of service. A peer that participates in a coalition lends ”bandwidth” to other peers of the coalition, in exchange for utility and consequently far greater download bandwidth. Simulation results have shown the effectiveness of the mechanism in stopping the free-riding peers and encouraging cooperation, increasing the performance of a P2P network and obtaining an improvement in time download performance.
Download
Paper Nr: 71
Title:

SHARED UNDERSTANDING AND SYNCHRONY EMERGENCE - Synchrony as an Indice of the Exchange of Meaning between Dialog Partners
Authors:

Ken Prepin and Catherine Pelachaud
Abstract: Synchrony is claimed by psychology as a crucial parameter of any social interaction. In dialog interactions, the synchrony between non-verbal behaviours of interactants is claimed to account for the quality of the interaction: to give to human a feeling of natural interaction, an agent must be able to synchronise on appropriate time. The synchronisation occurring during non-verbal iteractions has recently been modelised as a phenonomenon emerging from the coupling between interactants. We propose here, and test in simulation, a dynamical model of verbal communication which links the emergence of synchrony between non-verbal behaviours to the level of meaning exchanged through words by interactants: if partners of a dyad understand each other, synchrony emerges, whereas if they do not understand, synchrony is disrupted. In addition to retrieve the fact that synchrony emergence within a dyad of agents depends on their level of shared understanding, our tests pointed two noteworthy properties of synchronisation phenomenons: first, as well as synchrony accounts for mutual understanding and good interaction, di-synchrony accounts for misunderstanding; second, synchronisation and di-synchronisation emerging from mutual understanding are very quick phenomenons.
Download
Paper Nr: 74
Title:

A GAME THEORETIC BIDDING AGENT FOR THE AD AUCTION GAME
Authors:

Yevgeniy Vorobeychik
Abstract: TAC/AA (ad auction game) provides a forum for research into strategic bidding in keyword auctions to try out their ideas in an independently simulated setting. We describe an agent that successfully competed in the TAC/AA game, showing in the process how to operationalize game theoretic analysis to develop a very simple, yet highly competent agent. Specifically, we use simulation-based game theory to approximate equilibria in a restricted bidding strategy space, assess their robustness in a normative sense, and argue for relative plausibility of equilibria based on an analogy to a common agent design methodology. Finally, we offer some evidence for the efficacy of equilibrium predictions based on TAC/AA tournament data.
Download
Paper Nr: 91
Title:

STABILITY IN MATCHING PROBLEMS WITH WEIGHTED PREFERENCES
Authors:

Maria Silvia Pini, Francesca Rossi, Kristen Brent Venable and Toby Walsh
Abstract: The stable marriage problem is a well-known problem of matching men to women so that no man and woman, who are not married to each other, both prefer each other. Such a problem has a wide variety of practical applications, ranging from matching resident doctors to hospitals, to matching students to schools or more generally to any two-sided market. In the classical stable marriage problem, both men and women express a strict preference order over the members of the other sex, in a qualitative way. Here we consider stable marriage problems with weighted preferences: each man (resp., woman) provides a score for each woman (resp., man). Such problems are more expressive than the classical stable marriage problems. Moreover, in some real-life situations it is more natural to express scores (to model, for example, profits or costs) rather than a qualitative preference ordering. In this context, we define new notions of stability and optimality, and we provide algorithms to find marriages which are stable and/or optimal according to these notions. While expressivity greatly increases by adopting weighted preferences, we show that in most cases the desired solutions can be found by adapting existing algorithms for the classical stable marriage problem.
Download
Paper Nr: 107
Title:

CONTINUOUS PREFERENCES FOR ACTION SELECTION
Authors:

Emmanuelle Grislin-Le Strugeon and Patricia Everaere
Abstract: We have investigated the use of continuous alternatives for action selection by a behavior-oriented agent. Such an agent is made of concurrent ``behaviors"; each of these behaviors reacts to specific stimuli and provides a response according to a low-level goal. Since the behaviors are specialized, they can provide concurrent responses and conflicts among them must be solved to perform a coherent global behavior of the agent. In this context, voting methods allow to select only one of the responses of the behaviors, while taking into account their preferences and respecting all of their constraints. Previous works are based on action spaces limited to few discrete values and have shown difficulties in determining the behaviors weights for the vote. Furthermore, these works generally not allow to express the indifference of a behavior on a alternative's component, i.e. the fact that a behavior has no preference on the value of one component of an alternative. We propose in this article a method to use continuous values for the alternatives and a fair vote based on one alternative proposition per behavior. Our framework also allows the expression of indifference between alternatives. This proposition has been tested and compared, and the results show that our approach is better than previous propositions to avoid locked situations.
Download
Paper Nr: 129
Title:

CONDITIONAL GAME THEORY - A Generalization of Game Theory for Cooperative Multiagent Systems
Authors:

Wynn C. Stirling
Abstract: Game theory provides a framework within which to model multiagent systems. The conventional neoclassical theory is well suited for competitive scenarios where self-interest is the dominant concept of rational behavior, but is less appropriate for scenarios where opportunities for such complex social behavior as cooperation, compromise, and unselfishness are significant. Conditional game theory is an extension of the conventional neoclassical theory that permits agents to extend their spheres of interest beyond the self and enables them to condition their preferences on the preferences of other agents, thereby providing a mechanism with which to characterize complex social behavior. As these conditional preferences propagate through the system, social bonds are created among the players that permit notions of both group and individual preferences to emerge and, hence, for concepts of both group rationality and individual rationality to coexist. Computational complexity can often be mitigated by exploiting the sparseness of influence relationships among the members of the system.
Download
Paper Nr: 132
Title:

KEY POINTS FOR REALISTIC AGENT-BASED FINANCIAL MARKET SIMULATIONS
Authors:

Iryna Veryzhenko, Philippe Mathieu and Olivier Brandouy
Abstract: The purpose of this paper is to define software engineering abstractions that provide a generic framework for stock market simulations. We demonstrate a series of key points and principles that has governed the development of an Agent-Based financial market in the form of an API. The simulator architecture is presented. During artificial market construction we have faced the whole variety of agent-based modeling issues and solved them : local interaction, distributed knowledge and resources, heterogeneous environments, agents autonomy, artificial intelligence, speech acts, discrete scheduling and simulation. Our study demonstrates that the choices made for agent-based modeling in this context deeply impact the resulting market dynamics and proposes a series of advances regarding the main limits the existing platforms actually meet.
Download
Paper Nr: 135
Title:

FORMALIZING VIRTUAL ORGANIZATIONS
Authors:

Sergio Esparcia and Estefanía Argente
Abstract: This work presents a formalization of Virtual Organizations, which are designed by means of their structural entities, such as roles, organizational units or norms, and the dynamic entities that change through time like agents and groups. Entities are grouped by means of the Organizational Dimensions, explicitly represented in the proposed formalization. Additionally, a study of existing formalizations of Organization Centered Multiagent Systems is presented.
Download
Paper Nr: 147
Title:

SELF-ORGANIZING SYNCHRONICITY AND DESYNCHRONICITY USING REINFORCEMENT LEARNING
Authors:

Mihail Mihaylov, Yann-Aël Le Borgne, Ann Nowé and Karl Tuyls
Abstract: We present a self-organizing reinforcement learning (RL) approach for coordinating the wake-up cycles of nodes in a wireless sensor network in a decentralized manner. To the best of our knowledge we are the first to demonstrate how global synchronicity and desynchronicity can emerge through local interactions alone without the need of central mediator or any form of explicit coordination. We apply this RL approach to wireless sensor nodes arranged in different topologies and study how agents, starting with a random policy, are able to self-adapt their behavior based only on their interaction with neighboring nodes. Each agent independently learns to which nodes it should synchronize to improve message throughput and at the same with whom to desynchronize in order to reduce communication interference. The obtained results show how simple and computationally bounded sensor nodes are able to coordinate their wake-up cycles in a distributed way in order to improve the global system performance through (de)synchronicity.
Download
Paper Nr: 150
Title:

SELF-ORGANIZING SUPPLY NETWORKS - Autonomous Agent Coordination based on Expectations
Authors:

Jan Ole Berndt
Abstract: Supply networks are faced with the contradictory requirements of achieving high operational eciency while retaining the ability to adapt to a changing environment. Decentralized approaches representing logistics entities by autonomous artificial agents must therefore be enabled to structure and operate supply networks efficiently according to the domain’s inherent dynamics caused, for instance, by changing customer demands and network participants entering or leaving the system. In this paper, a novel approach to self- organization for multiagent systems is presented, avoiding a priori assumptions of agent characteristics by generating expectations from observable behavior.
Download
Paper Nr: 174
Title:

SAMPLING AND UPDATING HIGHER ORDER BELIEFS IN DECISION-THEORETIC BARGAINING WITH FINITE INTERACTIVE EPISTEMOLOGIES
Authors:

Paul Varkey and Piotr Gmytrasiewicz
Abstract: In this paper we study the sequential strategic interactive setting of bilateral, two-stage, seller-offers bargaining under uncertainty. We model the epistemology of the problem in a finite interactive decision-theoretic framework and solve it for three types of agents of successively increasing (epistemological) sophistication (i.e. capacity to represent and reason with higher orders of beliefs). We relax typical common knowledge assumptions, which, if made, would be sufficient to imply the existence of a, possibly unique, game-theoretic equilibrium solution. We observe and characterize a systematic monotonic relationship between an agent's beliefs and optimal behavior under a particular moment-based ordering of its beliefs. Based on this characterization, we present the \emph{spread-accumulate} technique of sampling an agent's higher order belief by generating ``evenly dispersed" beliefs for which we (pre)compute offline solutions. Higher order prior belief identification is then approximated to arbitrary precision by identifying a (previously solved) belief ``closest" to the true belief. These methods immediately suggest a mechanism for achieving a balance between efficiency and the quality of the approximation -- either by generating a large number of offline solutions or by allowing the agent to search online for a ``closer" belief in the vicinity of best current solution.
Download
Paper Nr: 177
Title:

ANNEXATIONS AND MERGING IN WEIGHTED VOTING GAMES - The Extent of Susceptibility of Power Indices
Authors:

Ramoni O. Lasisi and Vicki H. Allan
Abstract: This paper discusses weighted voting games and two methods of manipulating those games, called annexation and merging. These manipulations allow either an agent, called an annexer to take over the voting weights of some other agents in the game, or the coming together of some agents to form a bloc of manipulators to have more power over the outcomes of the games. We evaluate the extent of susceptibility to these manipulations in weighted voting games of the following prominent power indices: Shapley-Shubik, Banzhaf, and Deegan-Packel indices. We found that for unanimity weighted voting games of n agents and for the three indices: the manipulability, (i.e., the extent of susceptibility to manipulation) via annexation of any one index does not dominate that of other indices, and the upper bound on the extent to which an annexer may gain while annexing other agents is at most n times the power of the agent in the original game. Experiments on non unanimity weighted voting games suggest that the three indices are highly susceptible to manipulation via annexation while they are less susceptible to manipulation via merging. In both annexation and merging, the Shapley-Shubik index is the most susceptible to manipulation among the indices.
Download
Paper Nr: 292
Title:

COMPLETE DISTRIBUTED CONSEQUENCE FINDING WITH MESSAGE PASSING
Authors:

Katsumi Inoue, Gauvain Bourgne and Takayuki Okamoto
Abstract: When knowledge is physically distributed, information and knowledge of individual agents may not be collected to one agent because they should not be known to others for security and privacy reasons. We thus assume the situation that individual agents cooperate with each other to find useful information from a distributed system to which they belong, without supposing any master or mediate agent who collects all necessary information from the agents. Then we propose two complete algorithms for distributed consequence finding. The first one extends a technique of theorem proving in partition-based knowledge bases. The second one is a more cooperative method than the first one. We compare these two methods on a sample problem showing that both can improve efficiency over a centrlized approach, and then discuss other related approaches in the literature.
Download
Short Papers
Paper Nr: 19
Title:

INTENTIONAL MOBILE AGENTS IN UBIQUITOUS SYSTEMS
Authors:

Milene Serrano and Carlos José Pereira de Lucena
Abstract: Being everywhere, going anywhere and accessing at any time. Ubiquitous computing is the paradigm of service omnipresence, device heterogeneity, calm technology application and user satisfaction. Therefore, the success of ubiquitous systems depends on the mobile computing nature. In this paper, we introduce the application of intentional mobile agents in the systematic development of ubiquitous systems. These agents are commonly used to perform specific activities in dedicated servers, such as the content adaptability based on the ubiquitous profiles information. It demands context-awareness, which can be improved by exploring critical interactions among mobility, smart-spaces and cognitive-based autonomous entities. Finally, we show how our proposal has been appropriately applied to a ubiquitous system from the e-commerce domain.
Download
Paper Nr: 51
Title:

BEHAVIOR OF HOME CARE INTELLIGENT VIRTUAL AGENT WITH PRE-ThINK ARCHITECTURE
Authors:

Dilyana Budakova
Abstract: This paper considers the architecture and the behaviour of an intelligent virtual agent, taking care of the cosiness and the health-related features of a family house. The PRE-ThINK architecture is proposed and its components are considered. The dynamics of the decision making process in problem situations arising with the implementation of this architecture is shown. It is assumed that an agent, capable of taking the best possible decision in a critical situation will win the family members’ trust.
Download
Paper Nr: 65
Title:

CONVERGENCE ANALYSIS OF A MULTIAGENT COOPERATION MODEL
Authors:

Markus Eberling and Hans Kleine Büning
Abstract: Cooperation between autonomous and rational agents is still a challenge. The problem even gets harder if the agents follow different policies or if they are designed by different companies that have contradicting goals. In such systems agents cannot rely on the cooperation willingness of the other agents. Mostly, the reason for receiving cooperation is not observable as it is a result of the private decision process of the other agent. We deal with a multiagent system where the agents decide with whom to cooperate on the basis of multiple criteria. The system models these criteria with the help of rated propositions. Interaction in our system can only occur between agents that are linked together in a network structure. The agents adapt their values to the best performing neighbor and rewire their connections if they have uncooperative neighbors. We will present an imitation-based learning mechanism and we will theoretically analyze the mechanism. This paper also presents a worst case scenario in which the mechanism will fail.
Download
Paper Nr: 87
Title:

DETERMINING COOPERATION IN MULTIAGENT SYSTEMS WITH CULTURAL TRAITS
Authors:

Stefan Heinrich, Stefan Wermter and Markus Eberling
Abstract: Achieving cooperation among autonomous and rational agents is still a major challenge. In the past, altruistic cooperation was generally explained through genetic kinship relations. However, the theory of 'cultural kin' is an approach that tries to explain altruism through cultural relatedness. To promote cooperation among autonomous and rational agents, this work transfers the idea of cultural characteristics, which benefits social behaviour, to multiagent systems (MAS). Accordingly, agents are characterised by cultural traits, which they can imitate from their neighbours and are supposed to solve tasks, for which they need the cooperation of other agents in most cases. The interaction of cooperation and cultural trait propagation will be investigated in a theoretical analysis and in an empirical simulation in a particular developed framework. As a novelty, schemata will be analysed that are beyond the well-studied one-to-one interaction.
Download
Paper Nr: 95
Title:

ADAPTIVE STATE REPRESENTATIONS FOR MULTI-AGENT REINFORCEMENT LEARNING
Authors:

Yann-Michaël De Hauwere, Peter Vrancx and Ann Nowé
Abstract: When multiple agents act in the same environment, single-agent reinforcement learning (RL) techniques often fail, as they do not take into account other agents. An agent using single agent RL generally does not have sufficient information to obtain a good policy. However, multi-agent techniques that simply extend the state space to include information on the other agents suffer from a large overhead, leading to very slow learning. In this paper we describe a multi-level RL algorithm which acts independently whenever possible and learns in which states it should enrich its state information with information about other agents. Such states, which we call conflict states are detected using statistical information about expected payoffs in these states. We demonstrate through experiments that our approach learns a good trade-off between learning in the single-agent state space and learning in the multi-agent state space.
Download
Paper Nr: 116
Title:

AGENT BASED FRAMEWORK TO SIMULATE INHABITANTS’ BEHAVIOUR IN DOMESTIC SETTINGS FOR ENERGY MANAGEMENT
Authors:

Ayesha Kashif, Xuan Hoa Binh Le, Julie Dugdale and Stéphane Ploix
Abstract: Inhabitants' behaviour is a significant factor that influences energy consumption and has been previously incorporated as static activity profiles within simulation for energy control & management. In this paper an agent-based approach to simulate reactive/deliberative group behaviour has been proposed and implemented. It takes into account perceptual, psychological (cognitive), social behavioural elements and domestic context to generate reactive/deliberative behavioural profiles. The Brahms language is used to implement the proposed approach to learn behavioural patterns for energy control and management strategies.
Download
Paper Nr: 118
Title:

REGION-BASED HEURISTICS FOR AN ITERATIVE PARTITIONING PROBLEM IN MULTIAGENT SYSTEMS
Authors:

Thomas Kemmerich and Hans Kleine Büning
Abstract: Load balancing or access point selection in wireless networks both are problems where a large set of particles repeatedly has to be partitioned on another set of objects. In general this partitioning problem involves multiple contrary objectives. Due to the large number of particles a decentralized approach should be favored. In this work, such an iterative multi-objective optimization problem is modeled as multiagent system. We propose a local solution technique based on regions and some special coordination media. Agents select target objects based on the region they are in. Different region types are considered and a local heuristic is developed. We show the general potential of regions and experimentally analyze different approaches. All approaches are able to provide high quality solutions.
Download
Paper Nr: 142
Title:

RESOURCE ALLOCATION PROBLEMS ON NETWORKS - Maximizing Social Welfare using an Agent-based Approach
Authors:

Antoine Nongaillard and Philippe Mathieu
Abstract: Numerous applications can be formulated as an instance of resource allocation problems. Different kinds of solving techniques have been investigated, but the theoretical results cannot always be applied in practice due to inappropriate assumptions. Indeed, in these studies, agents are most of the time omniscient and/or have complete communication abilities. These hypotheses are not satisfied real life applications. practice. We propose in this paper a distributed mechanism leading to optimal solutions with respect to a more realistic environment. Agents only have limited perceptions and knowledge. Using local negotiations, they elaborate themselves optimal allocations, which can be viewed as emergent phenomena. We show that negotiations between individually rational agents lead to sub-optimal states in the society, and we propose a more suitable decision-making criterion, the sociability, leading to socially optimal solutions. Our method provides a sequence of transactions leading to optimal allocations, according to any communication networks, when four different welfare objectives are considered.
Download
Paper Nr: 155
Title:

THE COMPLEXITY OF MANIPULATING κ-APPROVAL ELECTIONS
Authors:

Andrew Lin
Abstract: An important problem in computational social choice theory is the complexity of undesirable behavior among agents, such as control, manipulation, and bribery in election systems, which are tempting at the individual level but disastrous for the agents as a whole. Creating election systems where the determination of such strategies is difficult is thus an important goal. An interesting set of elections is that of scoring protocols. Previous work in this area has demonstrated the complexity of misuse in cases involving a fixed number of candidates, and of specific election systems on unbounded number of candidates such as Borda. In contrast, we take the first step in generalizing the results of computational complexity of election misuse to cases of infinitely many scoring protocols on an unbounded number of candidates. We demonstrate the worst-case complexity of various problems in this area, by showing they are either polynomial-time computable, NP-hard, or polynomial-time equivalent to another problem of interest. We also demonstrate a surprising connection between manipulation in election systems and some graph theory problems.
Download
Paper Nr: 173
Title:

RESOURCE BOUNDED DECISION-THEORETIC BARGAINING WITH FINITE INTERACTIVE EPISTEMOLOGIES
Authors:

Paul Varkey and Piotr Gmytrasiewicz
Abstract: In this paper, we study the problem of bilateral bargaining under uncertainty. The problem is cast in an interactive decision-theoretic framework, in which the seller and the buyer agents are equipped with the ability to represent and reason with (probabilistic) beliefs about strategically relevant parameters, the other agent’s beliefs, the other agent’s beliefs about the current agent’s beliefs, and so on up to finite levels. The inescapable intractability of solving such models is characterized. We present a realization of the paradigm of (resource) bounded rationality by achieving a trade-off between optimality and efficiency as a function of the discretization resolution of the infinite action space. Memoization is used to further mitigate complexity and is realized here through disk-based caching. In addition, the inevitability of model extinction that arises in such settings is dealt with by indicating an intuitive realization of the absolute continuity condition based on maintaining an ensemble model, for e.g. a random model, that accounts for all actions not already accounted for by other models. Our results clearly demonstrate an operationalizable scheme for devising computationally efficient anytime algorithms on interactive decision-theoretic foundations for modeling (higher-order) epistemic dynamics and sequential decision making in multi agent domains with uncertainty.
Download
Paper Nr: 191
Title:

A MULTI-AGENT TOOL TO ANNOTATE BIOLOGICAL SEQUENCES
Authors:

Célia Ghedini Ralha, Hugo Wruck Schneider, Maria Emilia M. T. Walter and Marcelo M. Brígido
Abstract: Nowadays, great challenges are imposed by the existence of enormous volume of DNA and RNA sequences, which are continuously being discovered by genome sequencing projects, through the automatic sequencers based on massively parallel sequencing technologies. Thus, the task of identifying biological function for these sequences is a key activity in these high-throuput sequencing projects, where the automatic annotation must be significantly improved. In this context, this paper presents a multi-agent approach to address the important issue of automatic annotation in genome projects. We developed a sophisticated prototype named BioAgents, which simulates biologists knowledge and experience to annotate DNA or RNA sequences in genome sequencing projects, where different specialized intelligent agents work together to accomplish the annotation process.
Download
Paper Nr: 195
Title:

Norm-ML - A Modeling Language to Model Norms
Authors:

Karen Figueiredo and Viviane Torres da Silva
Abstract: Norms in multi-agent systems are mechanisms used to restrict the behavior of system entities during a period of time by defining what the entities are obligated, permitted or prohibited to do and by stating stimulus to their fulfillment by defining rewards and discouraging their violation by pointing out punishments. In this paper we propose a modeling language called NormML that makes possible the modeling of the norms together with its main properties and characteristics.
Download
Paper Nr: 199
Title:

AGENTS FOR MANAGING BUSINESS-TO-BUSINESS INTERACTIONS - Software Agents for Managing Business-to-Business Collaborations
Authors:

Edgar Tello-Leal, Omar Chiotti and Pablo D. Villarreal
Abstract: Current market opportunities and the growth of new Internet technologies encourage organizations to dynamically establish Business-to-Business (B2B) collaborations. B2B interactions are carried out by executing collaborative business processes among the parties. In this work we propose B2B collaboration agents for managing B2B interactions that allow organizations to dynamically establish collaborations and execute collaborative processes with their partners. The planning and execution of the actions of the agents that execute collaborative processes are driven by a Petri Net engine embedded in these agents. The role an organization fulfills in a collaborative process is represented by a high-level Petri Net model which is used to drive the behavior of the B2B collaboration agents representing the organization. Moreover, interaction protocols representing collaborative processes are executed by these agents without the need for protocols defined at design-time. Finally, an implementation of the B2B agents is presented.
Download
Paper Nr: 214
Title:

COOPERATIVE REPLIES TO UNBELIEVABLE ASSERTIONS - A Dialogue Protocol based on Logical Interpolation
Authors:

M. Nykänen, S. Eloranta, O. Niinivaara and R. Hakli
Abstract: We propose a dialogue protocol for situations in which an agent makes to another agent an assertion that the other agent finds impossible to believe. In this interaction, unbelievable assertions are rejected using explanations formed by logical interpolation and new assertions are being made such that all previous rebuttals are taken into account.
Download
Paper Nr: 233
Title:

A SCHIZOPHRENIC APPROACH FOR INTELLIGENT CONVERSATIONAL AGENTS
Authors:

Jean-Claude Heudin
Abstract: We present a novel approach for creating intelligent conversational agents based on a “schizophrenic” model implemented using the EVA (Evolutionary Virtual Agent) nano-agent architecture. The Ms House experiment developed using this approach is compared with Eliza and the Alice chatterbot.
Download
Paper Nr: 238
Title:

RAPID BEHAVIOUR MODELLING FOR AN AGENT-BASED SIMULATION

Authors:

Sascha A. Goldner
Abstract: Agent-based modelling and simulation has been applied to many different domains for studying highly complex systems. Usually these contain many different entities with their own specific behaviour patterns. The primary strength of agent-based simulation is to model and analyse human behaviour. In this context, one of the most complex and time-consuming tasks is the implementation of behavioural models for the human-like agents. In order to reduce this effort two additional methodologies are taken into consideration and applied to the agent-based model. Business process modelling and case-based reasoning is used for a rapid development of the behavioural part of an agent. This paper describes the scientific goals, ongoing work and interim results of the approach using the security system of an airport as an example.
Download
Paper Nr: 249
Title:

TRANSFER LEARNING FOR MULTI-AGENT COORDINATION
Authors:

Peter Vrancx, Yann-Michaël De Hauwere and Ann Nowé
Abstract: Transfer learning leverages an agent’s experience in a source task in order to improve its performance in a related target task. Recently, this technique has received attention in reinforcement learning settings. Training a reinforcement learning agent on a suitable source task allows the agent to reuse this experience to significantly improve performance on more complex target problems. Currently, reinforcement learning transfer approaches focus almost exclusively on speeding up learning in single agent systems. In this paper we investigate the potential of applying transfer learning to the problem of agent coordination in multi-agent systems. The idea underlying our approach is that agents can determine how to deal with the presence of other agents in a relatively simple training setting. By then generalizing this knowledge, the agents can use this experience to speed up learning in more complex multi-agent learning tasks.
Download
Paper Nr: 263
Title:

AGENT-BASED COMPUTER-GENERATED-FORCES’ BEHAVIOUR IMPROVEMENT
Authors:

Mike Bourassa, Nacer Abdellaoui and Glen Parkinson
Abstract: This paper captures the initial stages of a research project into improving the decision making performance of simulated entities in Computer Generated Forces (CGF) software applications. To date, the decisions made by Artificial Intelligence (AI)-enhanced synthetic entities have demonstrated a limited ability to react to changes in the synthetic environment, to use sensor data as effectively as a human operator, or in general to impact the synthetic environment in a comparable manner to a human operator. This paper presents a survey of AI in both the video gaming industry and academic circles leading to the proposal of a new agent architecture that combines a traditional agent architecture with a psychological framework (Maslow’s Hierarchy of Needs) leading to the specification of a “Needs-based” agent. This paper also captures the initial design decisions on the construction of a prototype and identifies candidate technologies to advance the research to the next phase. It is proposed that by combining the cognitive elements of the psychological framework with the behavioural emphasis of agents, synthetic entities in military and non-military simulations can produce better decisions and therefore exhibit more realistic behaviour which by ricochet will require less human intervention in simulation executions.
Download
Paper Nr: 264
Title:

PROBABILISTIC PLAN RECOGNITION FOR INTELLIGENT INFORMATION AGENTS - Towards Proactive Software Assistant Agents
Authors:

Jean Oh, Felipe Meneguzzi and Katia Sycara
Abstract: In this paper, we address probabilistic plan recognition techniques for a software assistant agent that can manage information on behalf of cognitively overloaded users, e.g., searching for necessary information regarding the user's current or future tasks and presenting information in the right format that is aligned with the user cognitive load. In this context, we present a flexible agent architecture for proactive information management, known here as ANTicipatory Information and Planning Agent (ANTIPA). We describe our plan prediction algorithm based on a decision-theoretic user model, and how the agent plans assistive actions for the predicted user plan. We describe a fully implemented agent of the ANTIPA architecture, and report preliminary user study results.
Download
Paper Nr: 273
Title:

AGENTS AND ONTOLOGIES FOR UNDERSTANDING AND PRESERVING THE ROCK ART OF MOUNT BEGO
Authors:

L. Papaleo, G. Quercini, V. Mascardi, M. Ancona, A. Traverso and H. De Lumley
Abstract: This paper describes the joint effort of computer scientists, archaeologists, and historians for designing a multi-agent system that exploits ontologies for the semantic description of the Mount Bego petroglyphs, thus moving a step forward their preservation. Most components of the MAS have already been developed and tested, and their integration is under way.
Download
Paper Nr: 274
Title:

INTEGRATED DYNAMICAL INTELLIGENCE FOR INTERACTIVE EMBODIED AGENTS
Authors:

Eric Aaron, Juan Pablo Mendoza and Henny Admoni
Abstract: For embodied agents that interact with people in time-sensitive applications, such as robot assistants or autonomous characters in video games, effectiveness can depend on responsive and adaptive behavior in dynamic environments. To support such behavior, agents' cognitive and physical systems can be modeled in a single, shared language of dynamical systems, an integrated design that supports performance with mechanisms not readily available in other modeling approaches. In this paper, we discuss these general ideas and describe how hybrid dynamical cognitive agents (HDCAs) employ such integrated modeling, resulting in dynamically sensitive user interaction, task sequencing, and adaptive behavior. We also present results of the first user-interactive applications of HDCAs: As demonstrations of this integrated cognitive-physical intelligence, we implemented our HDCAs as autonomous players in an interactive animated Tag game; resulting HDCA behavior included dynamic task re-sequencing, interesting and sensible unscripted behavior, and learning of a multi-faceted user-specified strategy for improving game play.
Download
Paper Nr: 276
Title:

DYNAMIC RESOURCE ALLOCATION THROUGH SEMI-STRUCTURED ADAPTATION
Authors:

Sander van Splunter, Elth Ogston and Frances Brazier
Abstract: Many of today’s systems are complex, distributed and networked, often situated in very dynamic environments. Such systems are often designed to adapt to change autonomically, to manage themselves autonomously. The Smart Energy Grid is an example of a large scale distributed system for which Distributed Energy Resource Management is crucial. This paper proposes a loosely coordinated management structure for Virtual Power Stations (VPS): hierarchical configuration. Within VPSs individual consumers and producers each with their own goals and responsibilities also share responsibility for collective goals such as reliability. Hierarchic self-management combines the strengths of centralised approaches with clear contracts and dependencies, with the strength of a fully decentralised approach within which distributed parts of a system adapt autonomously. Agent-based simulation experiments illustrate the potential of a hierarchical approach for distribution of resources within and between Virtual Power Stations as conditions change. Comparisons to centralised management and to fully decentralised management show that performance of the hierarchical approach is close to a centralised approach, whilst flexibility and scaleability are comparable to a fully decentralised approach.
Download
Paper Nr: 285
Title:

BLOCKING UNDERHAND ATTACKS BY HIDDEN COALITIONS
Authors:

Matteo Cristani, Erisa Karafili and Luca Viganò
Abstract: Similar to what happens between humans in the real world, in open multi-agent systems distributed over the Internet, such as online social networks or wiki technologies, agents often form coalitions by agreeing to act as a whole in order to achieve certain common goals. However, agent coalitions are not always a desirable feature of a system, as malicious or corrupt agents may collaborate in order to subvert or attack the system. In this paper, we consider the problem of hidden coalitions, whose existence and the purposes they aim to achieve are not known to the system, and which carry out underhand attacks, a term that we borrow from military terminology. We give a first approach to hidden coalitions by introducing a deterministic method that blocks the actions of potentially dangerous agents, i.e. possibly belonging to such coalitions. We also give a non-deterministic version of this method that blocks the smallest set of potentially dangerous agents. We calculate the computational cost of our two blocking methods, and prove their soundness and completeness.
Download
Paper Nr: 287
Title:

THE IMPORTANCE OF TIES IN THE EFFICIENCY OF CONVENTION EMERGENCE

Authors:

Paulo Urbano, João Balsa, Paulo Ferreira and João Baptista
Abstract: Social conventions are useful for the coordination of multi-agent systems. Decentralized models of social convention emergence have demonstrated that global agreement can be the result of local coordination behaviors without the need for any central control and authority. Convention arises through a co-learning process from repeated interactions, where the history of interactions plays a fundamental role in the learning process. The main research goal of this work is to study the role of ties in the standard frequency model called External Majority (EM). In the External Majority case agents change to a new convention only if a different convention was more often seen than the current one in the last μ interactions. Agents prefer to conserve their conventions if the current one is included in the set of the most often seen in the last μ encounters. We study three variations in EM behaviors regarding the way of dealing with tie situations and study empirically their impact on convention emergence efficiency. Efficiency is a decisive property in what concerns the design of large-scale self-organizing artificial systems, and one of the variations we propose strongly improves consensus emergence performance.
Download
Paper Nr: 327
Title:

THE EFFECTS OF MARKET DEMAND ON TRUTHFULNESS IN A COMPUTING RESOURCE OPTIONS MARKET

Authors:

Owen Rogers and Dave Cliff
Abstract: Grid, cluster and cloud computing provide the opportunity for computing resources to be traded as commodities in an open marketplace. An options market for computing resources would allow users to reserve a resource for a fee, and then pay an additional fee later should they actually need to use it. However, a major issue is ensuring that users do not falsify their likely requirements with the objective of reducing costs while keeping their right to use the resource. This paper describes an exploratory simulation implementation of a two-period model that was proposed by Wu, Zhang and Huberman (2008) which they claimed promoted truth-telling among the population of resource-buyers who interact with a Coordinator (a central vendor) of resources. Wu et al. provided a theoretical description and analysis of their model, but presented no empirical analysis of its commercial suitability. Our work, reported in this paper, explores the model's performance where demand for resources is variable and unpredictable. Using techniques similar to replicator dynamics (from studies of evolutionary processes in biology), we explore the behaviour of heterogeneous buyer populations under different market conditions. Through empirical and theoretical analysis, we determine the optimum honesty for which the Coordinator will most effectively prosper across a range of market conditions, and show how this data can be used to protect against risk.
Download
Paper Nr: 328
Title:

COOPERATION MECHANISM FOR A NETWORK GAME
Authors:

Alon Grubshtein and Amnon Meisels
Abstract: Many real world Multi Agent Systems encompass a large population of self interested agents which are connected with one another in an intricate network. If one is willing to accept the common axioms of Game Theory one can assume that the population will arrange itself into an equilibrium state. The present position paper proposes to use a mediating cooperative distributed algorithm instead. A setting where agents have to choose one action out of two - download information or free-ride their neighbors’ effort - has been studied recently. The present position paper proposes a method for constructing a Distributed Constraint Optimization Problem (DCOP) for a Network Game. The main result is that one can show that by cooperatively minimizing the constructed DCOP for a global solution all agents stand to gain at least as much as their equilibrium gain, and often more. This provides a mechanism for cooperation in a Network Game that is beneficial for all participating agents.
Download
Paper Nr: 394
Title:

OVERHEARING IN FINANCIAL MARKETS - A Multi-agent Approach
Authors:

Hedjazi Badiâa, Aknine Samir, Ahmed-Nacer Mohamed and Benatchba Karima
Abstract: Open complex systems as financial markets evolve in a highly dynamic and uncertain environment. They are often subject to significant fluctuations due to unanticipated behaviours and information. Modelling and simulating these systems by means of agent systems, i.e., through artificial markets is a valuable approach. In this article, we present our model of asynchronous artificial market consisting of a set of adaptive and heterogeneous agents in interaction. These agents represent the various market participants (investors and institutions). Investor Agents have advanced mental models for ordinary investors which do not relay on fundamental or technical analysis methods. On one hand, these models are based on the risk tolerance and on the other hand on the information gathered by the agents. This information results from overhearing influential investors in the market or the order books. We model the system through investor agents using learning classifier systems as reasoning models. As a result, our artificial market allows the study of overhearing impacts on the market. We also present the experimental evaluation results of our model.
Download
Paper Nr: 396
Title:

AGENT-HUMAN INTERACTIONS IN THE CONTINUOUS DOUBLE AUCTION, REDUX - Using the OpEx Lab-in-a-Box to explore ZIP and GDX
Authors:

Marco De Luca and Dave Cliff
Abstract: In 2001, a team of researchers at IBM published a paper in IJCAI which reported on the first experiments that systematically studied the interactions of human traders and software-agent traders in electronic marketplaces running the continuous double auction (CDA) mechanism. IBM found that two software-agent strategies, known as GD and ZIP, consistently outperformed human traders. IBM's results received international press coverage, probably because the CDA is the mechanism that is used in the main electronic trading systems that make up the global financial markets. In 2002, Tesauro & Bredin published details of an extension to GD, which they named GDX, for which they wrote: "We suggest that this algorithm may offer the best performance of any published CDA bidding strategy". To the best of our knowledge, GDX has never been tested against human traders under experimental conditions. In this paper, we report on the first such test: we present detailed analysis of the results from our own replications of IBM's human vs. ZIP experiments and from our world-first experiments that test humans vs. GDX. Our overall findings are that, both when competing against ZIP in pure agent vs. agent experiments and when competing against human traders, GDX's performance is significantly better than the performance of ZIP.
Download
Paper Nr: 408
Title:

“WHERE IS MY MIND”- THE EVOLUTION OF NPCS IN ONLINE WORLDS
Authors:

Magnus Johansson and Harko Verhagen
Abstract: Online worlds are complex places, where we have to know some of the rules of play to engage in the interaction. These worlds are both inhabited by human players and artificial agents called “non player characters” (NPCs). This is an article about how online worlds can contain a new level of interaction using more humanlike NPCs. We propose a new way to describe social interaction in online worlds, where NPCs are modelled to incorporate some of the traits that are more common to man. We also propose a way of analysing current NPCs and a way to create more humanlike NPCs that can contribute to a more unpredictable gaming experience, which seems to be the most promising aspect in the development of online worlds.
Download
Paper Nr: 21
Title:

RULE-BASED ORCHESTRATION OF AGENT-SOCIETIES
Authors:

Karl-Heinz Krempels and Christoph Terwelp
Abstract: Composing heterogeneous agent-based applications is mostly a complex task due to specific requirements of agents and existing dependencies among agents and societies. Resolving such dependency-networks is subject of agent and agent-society deployment and monitoring. The orchestration task covers automatic deployment, configuration, monitoring and reconfiguration of agent-based applications. Existing approaches provide static mapping of dependencies and constraints of agent and agent-society descriptions. This leads to a high modification effort, which requires very specialised developer’s know-how and can be very complex as well as error-prone, not only when distributing agents over several hosts, but also when launching agents locally. In this paper a reference model of a deployment infrastructure, a description model for agents and agent-societies and a knowledge-based mechanism for the orchestration of agent and agent-societies are presented with the aim to overcome the disadvantages of the considered existing approaches.
Download
Paper Nr: 50
Title:

SMART SOLUTIONS MULTI-AGENT PLATFORM FOR DYNAMIC TRANSPORTATION SCHEDULING

Authors:

Anton Ivaschenko, Alexander Tsarev, Alina Vaysblat and Petr Skobelev
Abstract: The paper presents an experience of Smart Solutions multi-agent platform development for real-time scheduling and optimization of mobile resources in transportation logistics which implements a concept of new generation of bio-inspired intelligent software. Main attention is given to representation of decision making strategies at different stages of the scheduling process.
Paper Nr: 72
Title:

INFLUENCE OF NEIGHBORHOOD AND SELF REORGANIZATION IN NETWORKED AGENTS
Authors:

Udara C. Weerakoon and Vicki H. Allan
Abstract: In a network graph in which nodes represent agents and edges represent "can work with" relationships, coalitions form. Such coalitions satisfy the skill set requirements of a task while still obeying partner requirements. Agents composing a coalition must form a connected subgraph in the network graph. There is no centralized control, and agents are free to propose any coalition that satisfies both the skill set and partner requirements. In this research, strengths of various coalition formation strategies are compared with respect to both success and profit. To determine the quality of the solution and for comparison purposes, we temporarily remove the restriction that an agent can belong to a single proposed coalition and that a task can be proposed by a single coalition (i.e. hedging environment). In addition, agents are given the ability to dynamically reorganize their partner connections in an attempt to improve utility. Agents employing egalitarian, intelligent and inventory reorganization are compared with agents employing structural and performance reorganization.
Download
Paper Nr: 84
Title:

A SOCIAL ROBOT FOR FACILITATING HUMAN RELATIONS IN SMART ENVIRONMENTS
Authors:

Berardina De Carolis, Nicole Novielli, Irene Mazzotta and Sebastiano Pizzutilo
Abstract: This paper describes how a robot may use social network analysis measures for facilitating social relations when acting as a “host” in a smart environment. The robot’s main goal consists in welcoming people, facilitating contacts and information sharing among people present in the environment. It uses knowledge on the structure of the social network for selecting the most appropriate strategy to create new relations or to spread information in the most effective way. To this aim, a multiagent system has been implemented for simulating and evaluating the functioning of the social facilitator.
Download
Paper Nr: 112
Title:

USING META-AGENTS TO BUILD MAS MIDDLEWARE
Authors:

S. C. Lynch
Abstract: Various multiagent platforms exist, each providing a range of individual capabilities but typically their implementations lack the flexibility to allow developers to adapt them to the differing needs of individual applications. This paper investigates the design of a kernel for MAS middleware based on primitive meta-agents. We specify these meta-agents and examine how they can be used to realise the capabilities required by multiagent platforms. We examine how changes in the organisation of meta-agents produce MAS platforms with differing behaviours. We evaluate the meta-agent approach by experimentation, demonstrating how modifications in meta-agent behaviour can provide different strategies for agent communication, scoping rules and connectivity with other tools.
Download
Paper Nr: 124
Title:

SCHEDULING BASED UPON FREQUENCY TRANSITION - Following Agents Agreement in a NCS
Authors:

O. Esquivel-Flores and H. Benitez-Pérez
Abstract: This paper provides a strategy to schedule a kind of real-time distributed system base upon changes on frequency transmission of agents included into a distributed system. Modifications on frequency transmission (sensing periods) of system’s individual components impact on system quality performance due to limited computing resources In this work we propose a dynamic linear time invariant model based upon frequency transmission and compute times of agent’s task which constitute a networked control system (NCS). Schedulability could be reached by controlling frequency transmission rates into a region bounded by minimum and maximum rates besides satisfy compute times. This idea is reinforced through a simulated case study based upon a helicopter simulation benchmark. It provides a good approximation of system response where main results are perform under a typical fault scenario for demonstration purposes.
Download
Paper Nr: 126
Title:

A NETWORK MODEL FOR PRICE STABILIZATION
Authors:

Jun Kiniwa and Kensaku Kikuta
Abstract: We consider a simple network model for economic agents where each can buy commodities in the neighborhood. Their prices may be initially distinct in any node. However, by assuming some rules on new prices, we show that the distinct prices will be converged to unique by iterating buy and sell operations. If we consider the price determination process as a kind of consensus problem, we can apply the stabilization proof to it. So we first present a naive protocol in which each agent always offers half of the difference between his own price and the lowest price in the neighborhood, called max price difference. Then, we consider game theoretic price determination in two ways, that is, by using different payoff functions. Finally, we propose a protocol in which each agent makes a bid uniformly distributed over the max price difference.
Download
Paper Nr: 151
Title:

TEAM FORMATION FOR AGENT COOPERATION IN LOGISTICS - Protocol Design and Complexity Analysis
Authors:

Arne Schuldt
Abstract: Supply network management is a challenging task due to the complexity, dynamics, and distribution of logistics processes. Delegating process control to intelligent software agents that represent logistics objects and act on their behalf helps approach these challenges. The resulting problem decomposition reduces the computational complexity. Dynamics can be dealt with locally. An important prerequisite for coordinated process control is that agents can cooperate with each other. Based on requirements from logistics, this paper presents an interaction protocol for team formation. A thorough complexity analysis for the proposed method is conducted because the arising interaction effort is not obvious as it depends on the number of teams formed. Therewith, agent developers can estimate the interaction effort and thus the applicability of the method in advance. Finally, an application of the introduced protocol is outlined.
Download
Paper Nr: 152
Title:

A GENOME BASED VISION OF MULTI-AGENT SYSTEMS
Authors:

Monica Vitali, Massimo Cossentino, Riccardo Rizzo and Salvatore Gaglio
Abstract: A set of software agents can be programmed to provide a large but finite set of services, often defined during design phase. After an evolution of the external environment, the pre-defined services could be unable to satisfy the requested quality. In this work an agent framework is proposed capable to adapt the agents in order to improve the quality of services provided by an agent society in correspondence with a modification of the external environment. These agents are based on a biologically inspired structure (genome), that defines all their behaviors and knowledges.
Download
Paper Nr: 162
Title:

CONDITIONS FOR LONG LASTING SUSTAINABLE INNOVATION IN AN AGENT-BASED MODEL
Authors:

Luca Ansaloni, Marco Villani, Roberto Serra and David Lane
Abstract: During the last decades, innovation has become a hot topic in a variety of socio-technological contexts: in particular, a key problem is that of understanding its origins. Moreover, scientists are not able to evaluate the sustainability of innovation processes, and it is difficult to discover what sort of conditions might lead to their crisis and even collapse. In this paper we present a model where agents are able to create new artifacts and can develop and enact strategies able to sustain innovation for very long periods. We discuss some results and make observations useful for understanding the processes and the strategies that sustain the growth of diversity in social and technological organizations.
Download
Paper Nr: 170
Title:

A GENERIC DECENTRALIZED RECURSIVE MULTIAGENT MODEL FOR MULTI-SCALE ORGANIZATION OF LARGE SCALE COMPLEX SYSTEMS
Authors:

Thi-Thanh-Ha Hoang, Michel Occello and Jean-Paul Jamont
Abstract: This paper proposes a generic multiagent decentralized architecture that is an answer to the multi-scale organization of large scale complex systems. Our proposal aims at supplying an application free multi-level recursive organization management, observation and integration software. The generic architecture is based on the AEIO decomposition and a recursive MAS model. A framework is developed on this model as a decentralized middleware which allows to real MAS to be communicated with agents in abstract layers.
Paper Nr: 171
Title:

MULTI-AGENT NEGOTIATION MODEL BASED-ON ARGUMENTATION IN THE CONTEXT OF E-COMMERCE
Authors:

Guorui Jiang, Yangwei Xu and Ying Liu
Abstract: In the e-commerce transactions, there are lots of commodities with the same name, but anyone of these commodities have certain attributes which differ itself from others. During the traditional process of multi-agent negotiation, only one commodity can be selected as the negotiation object from these commodities with same name, if buyer agent want to find an appropriate commodity, the flexibility and efficiency of multi-agent negotiation would be low. This paper studies the multi-agent negotiation model by argumentation for a group of commodities. It firstly defines all kinds of negotiation elements, then establishes a negotiation model based-on argumentation and describes the negotiation agreements and strategies, and finally an example would be presented for testifying the effects of this model.
Download
Paper Nr: 193
Title:

ATTENTION, MOTIVATION AND EMOTION IN COGNITIVE SOFTWARE AGENTS
Authors:

Daniela C. Terra, Henrique E. Borges and Paulo E. M. Almeida
Abstract: The observations that the emotional phenomenon is an essencial component to the living beings cognition has influenced the conception of artificial intelligent mechanisms. This influence has lead to the discussion whether it is possible to elaborate an intelligent system without including into it the emotions’s role. The proposed model implements an affective mechanism inside an architecture to build cognitive software agents, called ARTIFICE. Its conception was inspired in a biological bottom-up approach for classifying affections considering ideas and neuroscientific concepts about emotions and their influence on learning. Also it is considered that the attention, motivation and emotion are interdependent aspects that need to be considered together by an affective mechanism. The main objective of this work is to reproduce the adaptive functions of survival value in software. Moreover this study also aims to presents how the autonomy in artificial organisms can be acquired inserting appropriate synthetic emotions. The experiments suggest that the model is appropriate to allow agents to adapt themselves in generic environments, according to their incorporated affective structures. Their learning was accomplished solely from live interaction experiences with environment and other existing entities. No previous information about the artificial world were built-in into these agents.
Download
Paper Nr: 202
Title:

CONSTRUCTION OF EVOLUTIONARY MULTI-AGENT DOUBLE AUCTION MARKET FOR DATA MINING COMBINATIONAL STRATEGIES WITH STABLE RETURNS
Authors:

Chi Xu, Xiaoyu Zhao, Zheru Chi, Na Jia and Huiqun Zhao
Abstract: In this paper, we report a multi-agent artificial market scheme which evolves for imitating a real stock market as similar as better. The artificial market model can generate possible price trend curves those have high correlation coefficient with Hong Kong Hang Seng Index (HSI). The purpose of the imitation is to generate different possible market price dynamics, so the artificial markets can be reliable source to provide statistically analysis samples for a real market. The strategies optimized in the artificial markets can improve the stability and profitability in the real market. The aim of the experiment is to obtain a strategy that can provide profitable and relatively stable return without forecasting the future movements of the market.
Paper Nr: 207
Title:

USING DISTRIBUTED CSPs TO MODEL BUSINESS PROCESSES AGREEMENT IN SOFTWARE MULTIPROCESS
Authors:

Luisa Parody, María Teresa Gómez-López, Rafael M. Gasca and Diana Borrego
Abstract: A business process consists of a set of activities which are performed in a coordination way to obtain an objective. Sometimes the definition of this objective using only a classic business processes management is not possible. When the choreography of the processes cannot be defined with a combination of tasks using sequences, conditions, ’xor’, ’or’ and ’split’ control flow patterns, another representation and solution are necessary to be used. This problem makes difficult the decision making in software management projects. In this paper a way to describe a process agreement is described where the execution and the number of tasks execution order of the Web Services cannot be defined. As a case study, the resource distribution in a multiproject development environment is used. In this case, the processes have to achieve an agreement in function of the business rules that relate the processes. In order to achieve this objective, the Distributed Constraint Satisfaction Problems are used to model and solve this type of problems.
Download
Paper Nr: 209
Title:

E-HEALTH WORKFLOW-BASED AUTHORIZATION USING AN AGENT-ORIENTED VIRTUAL HEALTH CARE RECORD
Authors:

Andrei Vasilateanu and Luca D. Serbanati
Abstract: In this article a proposal to an integrated e-Health solution based on the Patient Electronic Health Record is presented. The main point is how the caregivers’ role that is obtained from authentification and authorization process is enforced in a cross-organizational care workflow using multi agent systems. Interoperability between healthcare organizations and provisioning of permission for accessing the medical record are also addressed using mediation and negotiation software agents. We envision the healthcare system as an open digital ecosystem, where multi-agent systems are organized in organizations.
Download
Paper Nr: 210
Title:

A MULTI-AGENT TRAFFIC SIMULATION FRAMEWORK FOR EVALUATING THE IMPACT OF TRAFFIC LIGHTS
Authors:

Raul Cajias, Antonio Gonzalez Pardo and David Camacho
Abstract: The growing of the number of vehicles cause serious strains on road infrastructures. Traffic jams inevitably occur, wasting time and money for both cities and their drivers. To mitigate this problem, traffic simulation tools based on multiagent techniques can be used to quickly prototype potentially problematic scenarios to better understand their inherent causes. This work centers around the effects of traffic light configuration on the flow of vehicles in a road network. To do so, a Multi-Agent Traffic Simulation Framework based on Particle Swarm Optimization techniques has been designed and implemented. Experimental results from this framework show an improvement in the average speed obtained by traffic controlled by adaptive over static traffic lights.
Download
Paper Nr: 225
Title:

A GENERAL APPROACH TO EXPLOIT ASPECTS OF INTELLIGENCE ON THE WEB

Authors:

Laura Burzagli and Francesco Gabbanini
Abstract: This contribution discusses the architecture of a software system that can be adopted to leverage the characteristics of Web 2.0 and Semantic Web, in order to make efficient usage of information. Key aspects on the implementation of a reusable framework are discussed, and the effectiveness of the approach is illustrated in an example scenario, in the context of inclusive e-Tourism.
Download
Paper Nr: 250
Title:

A SEMANTIC SCRAPING MODEL FOR WEB RESOURCES - Applying Linked Data to Web Page Screen Scraping
Authors:

José Ignacio Fernández-Villamor, Jacobo Blasco-García, Carlos Á. Iglesias and Mercedes Garijo
Abstract: In spite of the increasing presence of SemanticWeb Facilities, only a limited amount of the available resources in the Internet provide a semantic access. Recent initiatives such as the emerging Linked Data Web are providing semantic access to available data by porting existing resources to the semantic web using different technologies, such as database-semantic mapping and scraping. Nevertheless, existing scraping solutions are based on ad-hoc solutions complemented with graphical interfaces for speeding up the scraper development. This article proposes a generic framework for web scraping based on semantic technologies. This framework is structured in three levels: scraping services, semantic scraping model and syntactic scraping. The first level provides an interface to generic applications or intelligent agents for gathering information from the web at a high level. The second level defines a semantic RDF model of the scraping process, in order to provide a declarative approach to the scraping task. Finally, the third level provides an implementation of the RDF scraping model for specific technologies. The work has been validated in a scenario that illustrates its application to mashup technologies.
Download
Paper Nr: 256
Title:

MULTI-AGENT SYSTEMS IN DATA IMPUTATION OF COLLABORATIVE FILTERING - In Case of e-WeddingThailand

Authors:

Kunyanuth Kularbphettong, Phayung Meesad and Gareth Clayton
Abstract: Multi-agent system is rapidly emerging as a new paradigm to develop complex and intelligent commerce application systems in e Business. In this paper, we present the findings on the techniques used for data imputation techniques in Collaborative filtering based on multi-agent systems (MAS) of the on-going project, e-WeddingThailand. The aim of our project is to implement MAS combined with various techniques, like Web Services, Ontology, Web Semantic and Data Mining techniques. However, the present paper focuses on the data imputation technique in collaborative filtering utilized in order to treat missing values of customer behavioral patterns for Wedding business. As a result, a model obtained is therefore used as a benchmark for testing potential patterns so that they are used to strengthen the derived model in enhancing the overall system performance.
Download
Paper Nr: 282
Title:

LIVER TRANSPLANT WAITING LIST SIMULATION - An Agent based Model
Authors:

Alexander Flávio de Oliveira, Ricardo Poley Martins Ferreira and Agnaldo Soares de Lima
Abstract: Generally, Prioritizing is not a simple task. Given a waiting list for organ transplantation with dozens of patients, which patient must be prioritized at the time of an organ donation? Which patient has been waiting for longer time or which patient has the worst health? What policy would be fairer and more efficient? The search for an answer to this question can become a complex decision-making problem. The process for testing various policies and verifying whether each one of them can bring benefits or not, can be slow and consume valuable resources. Computer simulations can help by allowing, at a lower cost and with greater security and flexibility, reproduction and study of events whose real occurrence would not be desirable or even possible. These efforts result in the creation of tools for modeling and simulation. In this work, a model based on multi-agent systems was developed and implemented by using the Repast framework. The model was calibrated by using information taken from a real situation. Experiments were carried out illustrating situations in which the simulation model could be used. The results demonstrated the ability of the model to capture details of reality and to simulate defined situations.
Download
Paper Nr: 297
Title:

A TEAM-BASED ORGANIZATIONAL MODEL FOR ADAPTIVE MULTI-AGENT SYSTEMS

Authors:

Afsaneh Fatemi, Kamran Zamanifar, Naser Nemat bakhsh and Omid Askari
Abstract: Proper organizational modelling is a challenging issue in complex cooperative multi-agent systems. In this paper, we propose a team-based multi-agent organizational model, based on the Schwaninger's model of intelligent human organizations. It provides an integrative framework to rapid task handling, the main effectiveness requirement in many applications. Adaptation via reorganization makes the model suitable for dynamic, uncertain environments. Fast initial team formation, greedy capability-based coalition formation, and using the nearest neighbours’ resources improve utility compared to the identified hierarchical organizational models.
Download
Paper Nr: 374
Title:

CONTINUOUS ACTION REINFORCEMENT LEARNING AUTOMATA - Performance and Convergence
Authors:

Abdel Rodríguez, Ricardo Grau and Ann Nowé
Abstract: Reinforcement Learning is a powerful technique for agents to solve unknown Markovian Decision Processes, from the possibly delayed signals that they receive. Most RL work, in particular for multi-agent settings, assume a discrete action set. Learning automata are reinforcement learners, belonging to the category of policy iterators, that exhibit nice convergence properties in discrete action settings. Unfortunately, most applications assume continuous actions. A formulation for a continuous action reinforcement learning automaton already exists, but there is no convergence guarantee to optimal decisions. An improve of the performance of the method is proposed in this paper as well as the proof for the local convergence.
Download
Paper Nr: 382
Title:

HIVE-BDI: EXTENDING JASON WITH SHARED BELIEFS AND STIGMERGY
Authors:

Matteo Barbieri and Viviana Mascardi
Abstract: The classic BDI model focuses on the internal functioning of a single-agent architecture. Neither shared beliefs nor spontaneous and indirect coordination via the environment are supported. We describe Hive-BDI, an extension of the Jason BDI-style language with shared beliefs implemented via the logic-based coordination language ReSpecT and with stigmergy obtained via digital pheromones. A case study where robots roaming an unknown environment collaborate for creating a map demonstrates the feasibility of our approach.
Download
Paper Nr: 388
Title:

AN ASYNCHRONOUS MULTI-AGENT SYSTEM FOR OPTIMIZING SEMI-PARAMETRIC SPATIAL AUTOREGRESSIVE MODELS
Authors:

Matthias Koch and Tamás Krisztin
Abstract: Classical spatial autoregressive models share the same weakness as the classical linear regression models, namely it is not possible to estimate non-linear relationships between the dependent and independent variables. In the case of classical linear regression a semi-parametric approach can be used to address this issue. Therefore we propose an advanced semi-parametric modelling approach for spatial autoregressive models. Advanced semi-parametric modelling requires determining the best configuration of independent variable vectors, number of spline-knots and their positions. To solve this combinatorial optimization problem we propose an asynchronous multi-agent system based on genetic-algorithms. Three teams of agents work each on a subset of the problem and cooperate through sharing their most optimal solutions. Through this system we can derive more complex relationships, which are better suited for the often large and non-linear real-world problems faced by applied spatial econometricians.
Download
Paper Nr: 409
Title:

BRAINSTORMING - Agent based Meta-learning Approach
Authors:

Dariusz Plewczynski
Abstract: Brainstorming meta-learning approach is performed by a set of cognitive agents (CA), each implementing different machine learning (ML) algorithm, and/or trained using diverse subsets of available features describing input examples. The goal of the meta-learning procedure is providing a general and flexible classification meta-model for a given training data. In the first phase all agents, when trained using different features describing training objects, construct the ensemble of classification models independently. In the second step all solutions are gathered and the consensus is built between them by optimizing the voting weights for all agents. No early solution, given even by a generally low performing agent, is not discarded until the late phase of prediction, when comparing different learning models draws the final conclusion. The final phase, i.e. brainstorming tries to balance the generality of solution and the overall cognitive performance of all CAs. The classification meta-model is than used for predictions of the classification membership for given testing examples. The method was recently used in several ML applications in bioinformatics and chemoinformatics by the author.
Download