ICAART 2017 Abstracts

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

Area 1 - Artificial Intelligence

Full Papers
Paper Nr: 7
Title:

Deep Knowledge Representation based on Compositional Semantics for Chinese Geography
Authors:

Shengwu Xiong, Xiaodong Wang, Pengfei Duan, Zhe Yu and Abdelghani Dahou
Abstract: Elementary education resources for geography contain a wealth of knowledge that is a collection of information with various relationships. It is of vital importance to further develop human like intelligent technology for extracting deep semantic information to effectively understand the questions. In this paper, we propose a novel directed acyclic graph (DAG) deep knowledge representation built upon the theorem of combinational semantics. Knowledge is decomposed into nodes and edges which are then inserted into the ontology knowledge base. Experimental results demonstrate the superiority of the proposed method on question answering, especially when the syntax of question is complex, and its representation is fuzzy.
Download
Paper Nr: 8
Title:

Chinese Geographical Knowledge Entity Relation Extraction via Deep Neural Networks
Authors:

Shengwu Xiong, Jingjing Mao, Pengfei Duan and Shaohao Miao
Abstract: Aiming at the problem of complex relation pattern and low relation extraction precision in the unstructured free text, in this paper, a novel extraction model for Chinese geographical knowledge relation extraction using a real end-to-end deep neural networks (DNNs) is proposed. The proposed method is a fusion DNNs consisting of one convolutional neural networks and two neural networks, which contains word feature, sentence feature and class feature. For the experiments, we construct geographic entity relation type system and corpus. We achieve a good performance with the averaged overall precision of 96.54%, averaged recall of 92.99%, and averaged F value of 94.56%. Experimental results confirm the superiority of the proposed Chinese geographical knowledge relation extraction method. The data of this paper can be obtained from http://nlp.webmeteor.cn.
Download
Paper Nr: 10
Title:

Emotion Selection in a Multi-Personality Conversational Agent
Authors:

Jean-Claude Heudin
Abstract: Conversational agents and personal assistants represent an historical and important application field in artificial intelligence. This paper presents a novel approach to the problem of humanizing artificial characters by designing believable and unforgettable characters who exhibit various salient emotions in conversations. The proposed model is based on a multi-personality architecture where each agent implements a facet of its identity, each one with its own pattern of perceiving and interacting with the user. In this paper we focus on the emotion selection principle that chooses, from all the candidate responses, the one with the most appropriate emotional state. The experiment shows that a conversational multi-personality character with emotion selection performs better in terms of user engagement than a neutral mono-personality one.
Download
Paper Nr: 11
Title:

A Description Logic Based Knowledge Representation Model for Concept Understanding

Authors:

Farshad Badie
Abstract: This research employs Description Logics in order to focus on logical description and analysis of the phenomenon of ‘concept understanding’. The article will deal with a formal-semantic model for figuring out the underlying logical assumptions of ‘concept understanding’ in knowledge representation systems. In other words, it attempts to describe a theoretical model for concept understanding and to reflect the phenomenon of ‘concept understanding’ in terminological knowledge representation systems. Finally, it will design an ontology that schemes the structure of concept understanding based on the proposed semantic model.
Download
Paper Nr: 14
Title:

Paraconsistent Logic with Multiple Fuzzy Linguistic Truth-values
Authors:

Manren Wang and Xudong Luo
Abstract: This paper extends the two-valued paraconsistent logic into an one in which a proposition takes a truth-value from a set of multiple fuzzy linguistic terms. More specifically, we propose the corresponding inference rule and semantics, and finally prove the soundness of our new fuzzy logical system and its completeness. Moreover, we use an example to illustrate the applicability of our logic system in real life.
Download
Paper Nr: 23
Title:

Computing Maxmin Strategies in Extensive-form Zero-sum Games with Imperfect Recall
Authors:

Branislav Bosansky, Jiri Cermak, Karel Horak and Michal Pechoucek
Abstract: Extensive-form games with imperfect recall are an important game-theoretic model that allows a compact representation of strategies in dynamic strategic interactions. Practical use of imperfect recall games is limited due to negative theoretical results: a Nash equilibrium does not have to exist, computing maxmin strategies is NP-hard, and they may require irrational numbers. We present the first algorithm for approximating maxmin strategies in two-player zero-sum imperfect recall games without absentmindedness. We modify the well-known sequence-form linear program to model strategies in imperfect recall games resulting in a bilinear program and use a recent technique to approximate the bilinear terms. Our main algorithm is a branch-and-bound search that provably reaches the desired approximation after an exponential number of steps in the size of the game. Experimental evaluation shows that the proposed algorithm can approximate maxmin strategies of randomly generated imperfect recall games of sizes beyond toy-problems within few minutes.
Download
Paper Nr: 34
Title:

On using Support Vector Machines for the Detection and Quantification of Hand Eczema
Authors:

Stefan Schnürle, Marc Pouly, Tim vor der Brück, Alexander Navarini and Thomas Koller
Abstract: Hand eczema is one of the most frequent skin diseases affecting up to 14% of the population. Early detection and continuous observation of eczemas allows for efficient treatment and can therefore relieve symptoms. However, purely manual skin control is tedious and often error prone. Thus, an automatic approach that can assist the dermatologist with his work is desirable. Together with our industry partner swiss4ward, we devised an image processing method for hand eczema segmentation based on support vector machines and conducted several experiments with different feature sets. Our implementation is planned to be integrated into a clinical information system for operational use at University Hospital Zurich. Instead of focusing on a high accuracy like most existing state-of-the-art approaches, we selected F1 score as our primary measure. This decision had several implications regarding the design of our segmentation method, since all popular implementations of support vector machines aim for optimizing accuracy. Finally, we evaluated our system and achieved an F1 score of 58.6% for front sides of hands and 43.8% for back sides, which outperforms several state-of-the-art methods that were tested on our gold standard data set as well.
Download
Paper Nr: 36
Title:

Integration of Independence Detection into SAT-based Optimal Multi-Agent Path Finding - A Novel SAT-based Optimal MAPF Solver
Authors:

Pavel Surynek, Jiří Švancara, Ariel Felner and Eli Boyarski
Abstract: The problem of optimal multi-agent path finding (MAPF) is addressed in this paper. The task is to find optimal paths for mobile agents where each of them need to reach a unique goal position from the given start with respect to the given cost function. Agents must not collide with each other which is a source of combinatorial difficulty of the problem. An abstraction of the problem where discrete agents move in an undirected graph is usually adopted in the literature. Specifically, it is shown in this paper how to integrate independence detection (ID) technique developed for search based MAPF solving into a compilation-based technique that translates the instance of the MAPF problem into propositional satisfiability formalism (SAT). The independence detection technique allows decomposition of the instance consisting of a given number of agents into instances consisting of small groups of agents with no interaction across groups. These small instances can be solved independently and the solution of the original instance is combined from small solutions eventually. The reduction of the size of instances translated to the target SAT formalism has a significant impact on performance as shown in the presented experimental evaluation. The new solver integrating SAT translation and the independence detection is shown to be state-of-the-art in its class for optimal MAPF solving.
Download
Paper Nr: 39
Title:

Biologically-Inspired Neural Network for Walking Stabilization of Humanoid Robots
Authors:

Guilherme Barros Castro, Kazuya Tamura, Atsuo Kawamura and André Hirakawa
Abstract: In order to accomplish desired tasks, humanoid robots may have to deal with unpredicted disturbances, generated by objects, people and even ground imperfections. In some of these cases, foot placement is critical and cannot be changed. Furthermore, the robot has to conduct the actions planned meanwhile stabilizing its walking motion. Therefore, we propose a Biologically-inspired Neural Network (BiNN) to stabilize the walking motion of humanoid robots by ankle joint control, which minimally affects the current movements of the robot. In contrast to other neural networks, which only generate walking patterns, the BiNN is adaptive, as it compensates disturbances during the robot motion. Moreover, the BiNN has a low computational time and can be used as a module of other control methods. This approach was evaluated with Webots simulator, presenting improvements in the compensation of an external force in regard to its magnitude and duration.
Download
Paper Nr: 43
Title:

Assured Reinforcement Learning with Formally Verified Abstract Policies
Authors:

George Mason, Radu Calinescu, Daniel Kudenko and Alec Banks
Abstract: We present a new reinforcement learning (RL) approach that enables an autonomous agent to solve decision making problems under constraints. Our assured reinforcement learning approach models the uncertain environment as a high-level, abstract Markov decision process (AMDP), and uses probabilistic model checking to establish AMDP policies that satisfy a set of constraints defined in probabilistic temporal logic. These formally verified abstract policies are then used to restrict the RL agent's exploration of the solution space so as to avoid constraint violations. We validate our RL approach by using it to develop autonomous agents for a flag-collection navigation task and an assisted-living planning problem.
Download
Paper Nr: 47
Title:

Access Controlled Temporal Networks
Authors:

Carlo Combi, Roberto Posenato, Luca Viganò and Matteo Zavatteri
Abstract: We define Access-Controlled Temporal Networks (ACTNs) as an extension of Conditional Simple Temporal Networks with Uncertainty (CSTNUs). CSTNUs are able to handle features such as contingent durations and conditional constraints, and have thus been used to model the temporal constraints of workflows underlying business processes. However, CSTNUs are unable to model users and authorization constraints, and thus cannot model “who can do what, when”. ACTNs solve this problem by adding users and authorization constraints that must be considered together with temporal constraints. Dynamic controllability (DC) of ACTNs ensures the existence of an execution strategy, able to assign tasks to authorized users dynamically, satisfying all the relevant authorization constraints no matter what contingent durations turn out to be or what conditional constraints have to be considered. We show that the DC checking can be done via Timed Game Automata and provide experimental results using UPPAAL-TIGA on a concrete real-world case study.
Download
Paper Nr: 65
Title:

Extracting Contextonyms from Twitter for Stance Detection
Authors:

Guillaume Gadek, Josefin Betsholtz, Alexandre Pauchet, Stéphan Brunessaux, Nicolas Malandain and Laurent Vercouter
Abstract: Opinion mining on tweets is a challenge: short texts, implicit topics, inventive spellings and new vocabulary are the rule. We aim at efficiently determining the stance of tweets towards a given target. We propose a method using the concept of contextonyms and contextosets in order to disambiguate implicit content and improve a given stance classifier. Contextonymy is extracted from a word co-occurrence graph, and allows to grasp the sense of a word according to its surrounding words. We evaluate our method on a freely available annotated tweet corpus, used to benchmark stance detection on tweets during SemEval2016.
Download
Paper Nr: 67
Title:

Approximate Bayes Optimal Policy Search using Neural Networks
Authors:

Michael Castronovo, Vincent François-Lavet, Raphaël Fonteneau, Damien Ernst and Adrien Couëtoux
Abstract: Bayesian Reinforcement Learning (BRL) agents aim to maximise the expected collected rewards obtained when interacting with an unknown Markov Decision Process (MDP) while using some prior knowledge. State-of-the-art BRL agents rely on frequent updates of the belief on the MDP, as new observations of the environment are made. This offers theoretical guarantees to converge to an optimum, but is computationally intractable, even on small-scale problems. In this paper, we present a method that circumvents this issue by training a parametric policy able to recommend an action directly from raw observations. Artificial Neural Networks (ANNs) are used to represent this policy, and are trained on the trajectories sampled from the prior. The trained model is then used online, and is able to act on the real MDP at a very low computational cost. Our new algorithm shows strong empirical performance, on a wide range of test problems, and is robust to inaccuracies of the prior distribution.
Download
Paper Nr: 72
Title:

A Hierarchical Book Representation of Word Embeddings for Effective Semantic Clustering and Search
Authors:

Avi Bleiweiss
Abstract: Semantic word embeddings have shown to cluster in space based on linguistic similarities that are quantifiably captured using simple vector arithmetic. Recently, methods for learning distributed word vectors have progressively empowered neural language models to compute compositional vector representations for phrases of variable length. However, they remain limited in expressing more generic relatedness between instances of a larger and non-uniform sized body-of-text. In this work, we propose a formulation that combines a word vector set of variable cardinality to represent a verse or a sentence, with an iterative distance metric to evaluate similarity in pairs of non-conforming verse matrices. In contrast to baselines characterized by a bag of features, our model preserves word order and is more sustainable in performing semantic matching at any of a verse, chapter and book levels. Using our framework to train word vectors, we analyzed the clustering of bible books exploring multidimensional scaling for visualization, and experimented with book searches of both contiguous and out-of-order parts of verses. We report robust results that support our intuition for measuring book-to-book and verse-to-book similarity.
Download
Paper Nr: 76
Title:

Fast Many-to-One Voice Conversion using Autoencoders
Authors:

Yusuke Sekii, Ryohei Orihara, Keisuke Kojima, Yuichi Sei, Yasuyuki Tahara and Akihiko Ohsuga
Abstract: Most of voice conversion (VC) methods were dealing with a one-to-one VC issue and there were few studies that tackled many-to-one / many-to-many cases. It is difficult to prepare the training data for an application with the methods because they require a lot of parallel data. Furthermore, the length of time required to convert a speech by Deep Neural Network (DNN) gets longer than pre-DNN methods because the DNN-based methods use complicated networks. In this study, we propose a VC method using autoencoders in order to reduce the amount of the training data and to shorten the converting time. In the method, higher-order features are extracted from acoustic features of source speakers by an autoencoder trained with source speakers’ data. Then they are converted to higher-order features of a target speaker by DNN. The converted higher-order features are restored to the acoustic features by an autoencoder trained with data drawn from the target speaker. In the evaluation experiment, the proposed method outperforms the conventional VC methods that use Gaussian Mixture Models (GMM) and DNNs in both one-to-one conversion and many-to-one conversion with a small training set in terms of the conversion accuracy and the converting time.
Download
Paper Nr: 77
Title:

Japanese Text Classification by Character-level Deep ConvNets and Transfer Learning
Authors:

Minato Sato, Ryohei Orihara, Yuichi Sei, Yasuyuki Tahara and Akihiko Ohsuga
Abstract: Temporal (one-dimensional) Convolutional Neural Network (Temporal CNN, ConvNet) is an emergent technology for text understanding. The input for the ConvNets could be either a sequence of words or a sequence of characters. In the latter case there are no needs for natural language processing that depends on a language such as morphological analysis. Past studies showed that the character-level ConvNets worked well for news category classification and sentiment analysis / classification tasks in English and romanized Chinese text corpus. In this article we apply the character-level ConvNets to Japanese text understanding. We also attempt to reuse meaningful representations that are learned in the ConvNets from a large-scale dataset in the form of transfer learning, inspired by its success in the field of image recognition. As for the application to the news category classification and the sentiment analysis and classification tasks in Japanese text corpus, the ConvNets outperformed N-gram-based classifiers. In addition, our ConvNets transfer learning frameworks worked well for a task which is similar to one used for pre-training.
Download
Paper Nr: 96
Title:

Which Saliency Detection Method is the Best to Estimate the Human Attention for Adjective Noun Concepts?
Authors:

Marco Stricker, Syed Saqib Bukhari, Mohammad Al Naser, Saleh Mozafari, Damian Borth and Andreas Dengel
Abstract: This paper asks the question: how salient is human gaze for Adjective Noun Concepts (a.k.a Adjective Noun Pairs - ANPs)? In an existing work the authors presented the behavior of human gaze attention with respect to ANPs using eye-tracking setup, because such knowledge can help in developing a better sentiment classification system. However, in this work, only very few ANPs, out of thousands, were covered because of time consuming eye-tracking based data gathering mechanism. What if we need to gather the similar knowledge for a large number of ANPs? For example this could be required for designing a better ANP based sentiment classification system. In order to handle that objective automatically and without using an eye-tracking based setup, this work investigated if there are saliency detection methods capable of recreating the human gaze behavior for ANPs. For this purpose, we have examined ten different state-of-the-art saliency detection methods with respect to the ground-truths, which are human gaze pattern themselves over ANPs. We found very interesting and useful results that the Graph-Based Visual Saliency (GBVS) method can better estimate the human-gaze heatmaps over ANPs that are very close to human gaze pattern.
Download
Paper Nr: 99
Title:

SACAM - A Model for Describing and Classifying Sentiment Analysis Methods
Authors:

Aleksander Waloszek and Wojciech Waloszek
Abstract: In this paper we introduce SACAM — a model for describing and classifying sentiment analysis (SA) methods. The model focuses on the knowledge used during processing textual opinions. SACAM was designed to create informative descriptions of SA methods (or classes of SA methods) and is strongly integrated with its accompanying graphical notation suited for presenting the descriptions in diagrammatical form. The paper discusses applications of SACAM and shows directions of its further development.
Download
Paper Nr: 100
Title:

A Knowledge Driven Policy Framework for Internet of Things
Authors:

Emre Goynugur, Geeth De Mel, Murat Sensoy, Kartik Talamadupula and Seraphin Calo
Abstract: With the proliferation of technology, connected and interconnected devices (henceforth referred to as IoT) are fast becoming a viable option to automate the day-to-day interactions of users with their environment—be it manufacturing or home-care automation. However, with the explosion of IoT deployments we have observed in recent years, manually governing the interactions between humans-to-devices—and especially devices-to- devices—is an impractical task, if not an impossible task. This is because devices have their own obligations and prohibitions in context, and humans are not equip to maintain a bird’s-eye-view of the interaction space. Motivated by this observation, in this paper, we propose an end-to-end framework that (a) automatically dis- covers devices, and their associated services and capabilities w.r.t. an ontology; (b) supports representation of high-level—and expressive—user policies to govern the devices and services in the environment; (c) pro- vides efficient procedures to refine and reason about policies to automate the management of interactions; and (d) delegates similar capable devices to fulfill the interactions, when conflicts occur. We then present our initial work in instrumenting the framework and discuss its details.
Download
Paper Nr: 105
Title:

Where is that Button Again?! – Towards a Universal GUI Search Engine
Authors:

Sven Hertling, Markus Schröder, Christian Jilek and Andreas Dengel
Abstract: In feature-rich software a wide range of functionality is spread across various menus, dialog windows, toolbars etc. Remembering where to find each feature is usually very hard, especially if it is not regularly used. We therefore provide a GUI search engine which is universally applicable to a large number of applications. Besides giving an overview of related approaches, we describe three major problems we had to solve, which are analyzing the GUI, understanding the users’ query and executing a suitable solution to find a desired UI element. Based on a user study we evaluated our approach and showed that it is particularly useful if a not regularly used feature is searched for. We already identified much potential for further applications based on our approach.
Download
Paper Nr: 108
Title:

Assisted Feature Engineering and Feature Learning to Build Knowledge-based Agents for Arcade Games
Authors:

Bastian Andelefski and Stefan Schiffer
Abstract: Human knowledge can greatly increase the performance of autonomous agents. Leveraging this knowledge is sometimes neither straightforward nor easy. In this paper, we present an approach for assisted feature engineering and feature learning to build knowledge-based agents for three arcade games within the Arcade Learning Environment. While existing approaches mostly use model-free approaches we aim at creating a descriptive set of features for world modelling and building agents. To this end, we provide (visual) assistance in identifying and modelling features from RAM, we allow for learning features based on labeled game data, and we allow for creating basic agents using the above features. In our evaluation, we compare different methods to learn features from the RAM. We then compare several agents using different sets of manual and learned features with one another and with the state-of-the-art.
Download
Paper Nr: 110
Title:

Cluster Analysis of Twitter Data: A Review of Algorithms
Authors:

Noufa Alnajran, Keeley Crockett, David McLean and Annabel Latham
Abstract: Twitter, a microblogging online social network (OSN), has quickly gained prominence as it provides people with the opportunity to communicate and share posts and topics. Tremendous value lies in automated analysing and reasoning about such data in order to derive meaningful insights, which carries potential opportunities for businesses, users, and consumers. However, the sheer volume, noise, and dynamism of Twitter, imposes challenges that hinder the efficacy of observing clusters with high intra-cluster (i.e. minimum variance) and low inter-cluster similarities. This review focuses on research that has used various clustering algorithms to analyse Twitter data streams and identify hidden patterns in tweets where text is highly unstructured. This paper performs a comparative analysis on approaches of unsupervised learning in order to determine whether empirical findings support the enhancement of decision support and pattern recognition applications. A review of the literature identified 13 studies that implemented different clustering methods. A comparison including clustering methods, algorithms, number of clusters, dataset(s) size, distance measure, clustering features, evaluation methods, and results was conducted. The conclusion reports that the use of unsupervised learning in mining social media data has several weaknesses. Success criteria and future directions for research and practice to the research community are discussed.
Download
Paper Nr: 123
Title:

A Comparison between Asynchronous Backtracking Pseudocode and its JADEL Implementation
Authors:

Federico Bergenti, Eleonora Iotti, Stefania Monica and Agostino Poggi
Abstract: In this paper, a comparison between the pseudocode of a well-known algorithm for solving distributed constraint satisfaction problems and the implementation of such an algorithm in JADEL is given. First, background and motivations behind JADEL development are illustrated. Then, we make a description of the problem and a brief introduction to JADEL. The core of this work consists in the translation of the algorithm pseudocode in JADEL code, which is described in details. Scope of the paper is to evaluate such a translation, in terms of closeness to pseudocode, complexity, amount of code written and performance.
Download
Paper Nr: 125
Title:

Real-Time Data Harvesting Method for Czech Twitter
Authors:

Pavel Král and Václav Rajtmajer
Abstract: This paper deals with automatic analysis of Czech social media. The main goal is to propose an approach to harvest interesting messages from Twitter in Czech language with high download speed. This method uses user lists to discover potentially interesting tweets to download. It is motivated by the fact that only about 20% of Twitter users are posting informative messages, whereas the remaining 80% not and that it is possible to identify the "important" users by the user lists. The experimental results show that the proposed method is very efficient because it harvests about 6 times more data than the other approaches. This approach should be integrated into an experimental system for the Czech News Agency to monitor the current data-flow on Twitter, download messages in real-time, analyze them and extract relevant events.
Download
Paper Nr: 133
Title:

A Case Base Approach to Cardiovascular Diseases using Chest X-ray Image Analysis
Authors:

Ricardo Faria, Victor Alves, Filipa Ferraz, João Neves, Henrique Vicente and José Neves
Abstract: Cardio Vascular Disease (CVD) also known as heart and circulatory disease comprises all the illnesses of the heart and the circulatory system, namely coronary heart disease, angina, heart attack, congenital heart disease or stroke. CVDs are, nowadays, one of the main causes of death. Indeed, this fact reveals the centrality of prevention and how important is to be aware on these kind of situations. Thus, this work will focus on the development of a decision support system to help to prevent these events from happening, centred on a formal framework based on Mathematical Logic and Logic Programming for Knowledge Representation and Reasoning, complemented with a Case Based Reasoning approach to computing that caters to the handling of incomplete, unknown or even self-contradictory information or knowledge.
Download
Paper Nr: 138
Title:

Lifelong Machine Learning with Adaptive Multi-Agent Systems
Authors:

Nicolas Verstaevel, Jérémy Boes, Julien Nigon, Dorian d'Amico and Marie-Pierre Gleizes
Abstract: Sensors and actuators are progressively invading our everyday life as well as industrial processes. They form complex and pervasive systems usually called ”ambient systems” or ”cyber-physical systems”. These systems are supposed to efficiently perform various and dynamic tasks in an ever-changing environment. They need to be able to learn and to self-adapt throughout their life, because designers cannot specify a priori all the interactions and situations they will face. These are strong requirements that push the need for lifelong machine learning, where devices can learn models and behaviours during their whole lifetime and are able to transfer them to perform other tasks. This paper presents a multi-agent approach for lifelong machine learning.
Download
Paper Nr: 142
Title:

Tuning Agent's Profile for Similarity Measure in Description Logic ELH
Authors:

Teeradaj Racharak and Satoshi Tojo
Abstract: In Description Logics (DLs), concept similarity measure aims at identifying a degree of commonality of two given concepts and is often regarded as a generalization of the classical reasoning problem of equivalence. That is, any two concepts are equivalent if and only if their similarity degree is one. When two concepts are not equivalent, the level of similarity varies depending not only on the objective factors (e.g. the structure of concept descriptions) but also on the subjective factors (i.e. the agent’s preferences). Realistic ontologies are generally complex. Methodologies for tuning a measure to conform with the agent’s preferences should be practical, i.e. it is doable in practice. In this work, we investigate and formalize the task of tuning the preference functions based on the information defined in a TBox and an ABox. We also show how the proposed approaches can be reconciled with the measure sim π, i.e. a concept similarity measure under preference profile for DL ELH . Finally, the paper relates the approach to others and discusses future direction.
Download
Paper Nr: 147
Title:

Enhancing Pigeon-Hole based Encoding of Boolean Cardinality Constraints
Authors:

Soukaina Hattad, Said Jabbour, Lakhdar Sais and Yakoub Salhi
Abstract: In this paper, we propose to deal with the encoding of cardinality constraints ∑ni=1 xi ≥ b into conjunctive normal form. We consider the one proposed recently (Jabbour et al., 2014) based on pigeon-hole problem. Then, we show that even if the number of clauses of the CNF based encoding is in O(b x (n - b)),, the number of literals of resulting formula can be much more higher O(b(n-b)²)$. To decrease the complexity in terms of number of literals, we propose a compact representation of some clauses of the encoding. Our approach allows to have a quadratic encoding in terms of literals while maintaining the same complexity in terms of clauses and additional variables. An experimental evaluation is performed to show the competitiveness of the new encoding.
Download
Paper Nr: 151
Title:

Modelling and Reasoning with Uncertain Event-observations for Event Inference
Authors:

Sarah Calderwood, Kevin McAreavey, Weiru Liu and Jun Hong
Abstract: This paper presents an event modelling and reasoning framework where event-observations obtained from heterogeneous sources may be uncertain or incomplete, while sensors may be unreliable or in conflict. To address these issues we apply Dempster-Shafer (DS) theory to correctly model the event-observations so that they can be combined in a consistent way. Unfortunately, existing frameworks do not specify which event-observations should be selected to combine. Our framework provides a rule-based approach to ensure combination occurs on event-observations from multiple sources corresponding to the same event of an individual subject. In addition, our framework provides an inference rule set to infer higher level inferred events by reasoning over the uncertain event-observations as epistemic states using a formal language. Finally, we illustrate the usefulness of the framework using a sensor-based surveillance scenario.
Download
Paper Nr: 152
Title:

Concept and Realization of a Diagnostic System for Smart Environments
Authors:

Eric Heiden, Sebastian Bader and Thomas Kirste
Abstract: Automatically diagnosing a complex system containing heterogeneous hard- and software components is a challenging task. To analyse the problem, we first describe different scenarios a diagnostic engine might be confronted with. Based on those scenarios, a concept and an implementation of a semi-automatic diagnostic system are presented and some first benchmarks are shown.
Download
Paper Nr: 153
Title:

Long Range Optical Truck Tracking
Authors:

Christian Winkens and Dietrich Paulus
Abstract: Platooning applications require precise knowledge about position and orientation (pose) of the leading vehicle especially in rough terrain. We present an optical solution for a robust pose estimation using artificial markers and a camera as the only sensor. Temporal coherence of image sequences is used in a Kalman filter to obtain precise estimates. Furthermore based on the marker detections we utilize an adaptive model building algorithm which learns a keypoint based representation of the leading vehicle at runtime. The model is continuously updated and allows a markerless tracking of the vehicle for up to 70meters even when driving at high velocities. The system is designed for and tested in off-road scenarios. A pose evaluation is performed in a simulation testbed.
Download
Paper Nr: 155
Title:

Reasoning for Autonomous Agents in Dynamic Domains
Authors:

Stephan Opfer, Stefan Jakob and Kurt Geihs
Abstract: In contrast to simple autonomous vacuum cleaners, multi-purpose robots that fetch a cup of coffee and clean up rooms require cognitive skills such as learning, planning, and reasoning. Especially reasoning in dynamic and human populated environments demands for novel approaches that can handle comprehensive and fluent knowledge bases. A promising approach is Answer Set Programming (ASP), offering multi-shot solving techniques and non-monotonic stable model semantics. Our objective is to equip multi-agent systems with ASP-based reasoning capabilities, enabling a team of robots to cope with dynamic environments. Therefore, we combined ALICA - A Language for Interactive Cooperative Agents - with the ASP solver Clingo and chose topological path planning as our evaluation scenario. We utilised the Region Connection Calculus as underlying formalism of our evaluation and investigated the scalability of our implementation. The results show that our approach handles dynamic environments and scales up to appropriately large problem sizes.
Download
Short Papers
Paper Nr: 9
Title:

Operationalization of the Blending and the Levels of Abstraction Theories with the Timed Observations Theory
Authors:

Marc Le Goc and Fabien Vilar
Abstract: Providing a meaning to observations coming from humans (interviews) or machines (data sets) is a necessity to build adequate analysis and efficient models that can be used to take a decision in a given domain. Fauconnier and Turner demonstrates in 1998 the cognitive power of their Blending Theory where the blending of multiple conceptual networks is presented as a general-purpose, fundamental, indispensable cognitive operation to this aim. On the other hand, Floridi proposed in 2008 a theory of levels of abstraction as a fundamental epistemological method of conceptual analysis that can also be used to this aim. Both theories complete together but both lack of mathematical foundations to build an operational data and knowledge modeling method that helps and guides the Analysts and the Modeling Engineers. In this theoretical paper, we introduce the mathematical framework, based on the Timed Observations Theory, designed to build a method of abstraction merging together the Blending Theory and the Levels of Abstraction Theory. Up to our knowledge, this is the first mathematical theory allowing the operationalization of the Blending Theory and the Levels of Abstraction Theory. All over the paper, the mathematical framework is illustrated on an oral exchange between three persons observing a vehicle. We show that this framework allows to build a rational meaning of this exchange under the form of a superposition of three abstraction levels.
Download
Paper Nr: 12
Title:

Measuring the Performance of Push-ups - Qualitative Sport Activity Recognition
Authors:

Sebastian Baumbach and Andreas Dengel
Abstract: The trend of mobile activity monitoring using widely available technology is one of the most blooming concepts in the recent years. It supports many novel applications, such as fitness games or health monitoring. In these scenarios, activity recognition tries to distinguish between different types of activities. However, only little work has focused on qualitative recognition so far: How exactly is the activity carried out? In this paper, an approach for supervising activities, i.e. qualitative recognition, is proposed. The focus lied on push-ups as a proof of concept, for which sensor data of smartphones and smartwatches were collected. A user-dependent dataset with 4 participants and a user-independent dataset with 16 participants were created. The performance of Naive Bayes classifier was tested against normal, kernel and multivariate multinomial probability distributions. An accuracy of 90.5% was achieved on the user-dependent model, whereas the user-independent model scored with an accuracy of 80.3%.
Download
Paper Nr: 15
Title:

Novelty and Objective-based Neuroevolution of a Physical Robot Swarm
Authors:

Forrest Stonedahl, Susa H. Stonedahl, Nelly Cheboi, Danya Tazyeen and David Devore
Abstract: This paper compares the use of novelty search and objective-based evolution to discover motion controllers for an exploration task wherein mobile robots search for immobile targets inside a bounded polygonal region and stop to mark target locations. We evolved the robots' neural-network controllers in a custom 2-D simulator, selected the best performing neurocontrollers from both novelty search and objective-based search, and compared performance relative to an unevolved (baseline) controller and a simple human-designed controller. The controllers were also transferred onto physical robots, and the real-world tests provided good empirical agreement with simulation results, showing that both novelty search and objective-based search produced controllers that were comparable or superior to the human-designed controller, and that objective-based search slightly outperformed novelty search. The best controllers had surprisingly low genotypic complexity, suggesting that this task may lack the type of deceptive fitness landscape that has previously favored novelty search over objective-based search.
Download
Paper Nr: 20
Title:

Commonsense Reasoning in a Deeper Way: By Discovering Relations between Predicates
Authors:

Wenguan Huang and Xudong Luo
Abstract: One of the biggest drawbacks of nowadays AI reasoning systems is their lack of commonsense. To address the issue, some commonsense knowledge bases and a bunch of reasoning mechanisms with them have been developed to tackle this problem. However, most of them concentrate on the relation between entities (e.g., "cat" and "fish"), but few discuss the relation between predicates (e.g., "angry" and "shout"), which fall into a deeper level of commonsense. To the end, in this paper, we develop a commonsense reasoning framework, which focuses on this type of commonsense knowledge. More specifically, first we give a formal definition of this kind of commonsense. Then we construct a set of knowledge by extending the predicate set of ConceptNet, and apply information extraction technique to capture them from corpus. Finally, to evaluate our framework, we conduct experiments against a part of the Winograd Schema Challenge, which, its author claimed, is an alternative of Turing Test. The result of our experiments confirms the effectiveness of our framework.
Download
Paper Nr: 21
Title:

A Polynomial Algorithm for Merging Lightweight Ontologies in Possibility Theory Under Incommensurability Assumption
Authors:

Salem Benferhat, Zied Bouraoui, Ma Thi Chau, Sylvain Lagrue and Julien Rossit
Abstract: The context of this paper is the one of merging lightweight ontologies with prioritized or uncertain assertional bases issued from different sources. This is especially required when the assertions are provided by multiple and often conflicting sources having different reliability levels. We focus on the so-called egalitarian merging problem which aims to minimize the dissatisfaction degree of each individual source. The question addressed in this paper is how to merge prioritized assertional bases, in a possibility theory framework, when the uncertainty scales are not commensurable, namely when the sources do not share the same meaning of uncertainty scales. Using the notion of compatible scale, we provide a safe way to perform merging. The main result of the paper is that the egalitarian merging of prioritized assertional bases can be achieved in a polynomial time even if the uncertainty scales are not commensurable.
Download
Paper Nr: 31
Title:

Superposition of Qualitative Rectangles using a Quantitative Model
Authors:

Takeaki Kato, Sosuke Moriguchi and Kazuko Takahashi
Abstract: This paper describes an approach to qualitative problem-solving using the quantitative method in spatial reasoning. We consider the superposition of two objects, such that pre-specified parts of the objects are visible. First, we qualify an object to create a model. It is expressed as a matrix of tiles, which are either black or white depending on the visibility requirement. We use this to determine the location of two objects. This process involved quantitative treatment. We describe a sound and complete algorithm that provides quantitative solutions and implemented it as a system with a graphical user interface. Then, we extend this algorithm so that we may search for a better solution considering a qualitatively equivalent model of the objects; that is, the topological relationships between the black and white regions are identical. This approach is useful for analyzing or designing a projection of three-dimensional objects onto a two-dimensional plane, because it not only reduces the computational expense but also provides a better fit with common sense and human reasoning.
Download
Paper Nr: 33
Title:

Distribution Data Across Multiple Cloud Storage using Reinforcement Learning Method
Authors:

Abdullah Algarni and Daniel Kudenko
Abstract: Storing data on a single cloud storage service may cause several potential problems for the data owner such as service continuity, availability, performance, security, and the risk of vendor lock-in. A promising solution to tackle some of these issues is to distribute the data across multiple cloud storage services (MCSS). However, the distinguishing characteristics of different cloud providers, in terms of pricing schemes and service performance, make it difficult to optimise the cost and the performance concurrently on MCSS. This paper proposes a framework for automatically tuning the data distribution policies across MCSS from the client side based on file access patterns. The aim of this work is to optimise the average cost and the average service performance (mainly latency time) on MCSS. To achieve this goal, two different machine learning algorithms are used in this work: (1) supervised learning to predict file access patterns, and (2) reinforcement learning to control data distribution parameters based on the prediction of file access pattern. The framework was tested on a cloud storage emulator, where its was set to act like several common cloud storage services. The result of testing this framework shows a significant improvement in the cost and performance of storing data in multiple clouds, as compared to the commonly used uniform file distribution.
Download
Paper Nr: 46
Title:

New Flow-based Heuristic for Search Algorithms Solving Multi-agent Path Finding
Authors:

Jiri Svancara and Pavel Surynek
Abstract: We address the problem of optimal multi-agent path finding (MAPF) in this paper. The task is to find a set of actions for each agent in know terrain so that each agent arrives to its desired destination from a given starting position. Agents are not allowed to collide with each other along their paths. Furthermore, a solution that minimizes the total time is required. In this paper we study search-based algorithms that systematically explore state space. These algorithms require a good heuristic function that can improve the computational effectiveness by changing the order in which the states are expanded. We propose such new heuristic, which is based on relaxation of MAPF solving via its reduction to multi-commodity flow over time expanded graph. The multi-commodity flow is relaxed to single commodity flow, which can be solved in polynomial time. We show that our new heuristic is monotone and therefore can be used in search-based algorithms effectively. We also give theoretical analysis of the new heuristic and compare it experimentally with base-line heuristics that are often used.
Download
Paper Nr: 63
Title:

Dynamic Programming for One-sided Partially Observable Pursuit-evasion Games
Authors:

Karel Horák and Branislav Bošanský
Abstract: Pursuit-evasion scenarios appear widely in robotics, security domains, and many other real-world situations. We focus on two-player pursuit-evasion games with concurrent moves, infinite horizon, and discounted rewards. We assume that the players have partial observability, however, the evader has an advantage of knowing the current position of pursuer’s units. This setting is particularly interesting for security domains where a robust strategy, maximizing the utility in the worst-case scenario, is often desirable. We provide, to the best of our knowledge, the first algorithm that provably converges to the value of a partially observable pursuit-evasion game with infinite horizon. Our algorithm extends well-known value iteration algorithm by exploiting that (1) value functions of our game depend only on the position of the pursuer and the belief he has about the position of the evader, and (2) that these functions are piecewise linear and convex in the belief space.
Download
Paper Nr: 85
Title:

A Cognitive Approach for Reproducing the Homing Behaviour of Honey Bees
Authors:

Xin Yuan, Michael John Liebelt and Braden J. Phillips
Abstract: We describe the implementation if an agent-based controller for an autonomous robot with cognitive abilities that reproduce homing capability in the foraging behaviour of the honeybee. The agent is based on a symbolic representation of data and information and is written in a language designed to describe fine-grained large scale parallelism, the Street language (Frost et al. 2015). The objective of this approach is to enable the direct translation of agents written in Street into embedded hardware, to achieve compact, power efficient, autonomous cognitive processing capability.
Download
Paper Nr: 89
Title:

Adversarial Reinforcement Learning in a Cyber Security Simulation
Authors:

Richard Elderman, Leon J. J. Pater, Albert S. Thie, Madalina M. Drugan and Marco M. Wiering
Abstract: This paper focuses on cyber-security simulations in networks modeled as a Markov game with incomplete information and stochastic elements. The resulting game is an adversarial sequential decision making problem played with two agents, the attacker and defender. The two agents pit one reinforcement learning technique, like neural networks, Monte Carlo learning and Q-learning, against each other and examine their effectiveness against learning opponents. The results showed that Monte Carlo learning with the Softmax exploration strategy is most effective in performing the defender role and also for learning attacking strategies.
Download
Paper Nr: 91
Title:

An Evolutionary Traveling Salesman Approach for Multi-Robot Task Allocation
Authors:

Muhammad Usman Arif and Sajjad Haider
Abstract: Multi-Robot Task Allocation (MRTA) addresses the problems related to an efficient job assignment in a team of robots. This paper expresses MRTA as a generalization of the Multiple Traveling Salesman Problem (MTSP) and utilizes evolutionary algorithms (EA) for optimal task assignment. The MTSP version of the problem is also solved using combinatorial optimization techniques and results are compared to demonstrate that EA can be effectively used for providing solutions to such problems.
Download
Paper Nr: 92
Title:

Probabilistic Multi-Agent Plan Recognition as Planning (P-Maprap): Recognizing Teams, Goals, and Plans from Action Sequences
Authors:

Chris Argenta and Jon Doyle
Abstract: We extend Multi-agent Plan Recognition as Planning (MAPRAP) to Probabilistic MAPRAP (P-MAPRAP), which probabilistically identifies teams and their goals from limited observations of on-going individual agent actions and a description of actions and their effects. These methods do not rely on plan libraries, as such are infeasibly large and complex in multi-agent domains. Both MAPRAP and P-MAPRAP synthesize plans tailored to hypothesized team compositions and previous observations. Where MAPRAP prunes team-goal interpretations on optimality grounds, P-MAPRAP directs its search base on a likelihood ranking of interpretations, thus effectively reducing the synthesis effort needed without compromising recognition. We evaluate performance in scenarios that vary the number of teams, agent counts, initial states, goals, and observation errors, assuming equal base-rates. We measure accuracy, precision, and recall online to evaluate early stage recognition. Our results suggest that compared to MAPRAP, P-MAPRAP exhibits improved speed and recognition accuracy.
Download
Paper Nr: 97
Title:

Hierarchical Self-organizing Maps System for Action Classification
Authors:

Zahra Gharaee, Peter Gärdenfors and Magnus Johnsson
Abstract: We present a novel action recognition system that is able to learn how to recognize and classify actions. Our system employs a three-layered neural network hierarchy consisting of two self-organizing maps together with a supervised neural network for labelling the actions. The system is equipped with a module that pre-processes the 3D input data before the first layer, and a module that transforms the activity elicited over time in the first layer SOM into an ordered vector representation before the second layer, thus achieving a time invariant representation. We have evaluated our system in an experiment consisting of ten different actions selected from a publicly available data set with encouraging result.
Download
Paper Nr: 107
Title:

Keeping Secrets in Modalized DL Knowledge Bases
Authors:

Gopalakrishnan Krishnasamy Sivaprakasam and Giora Slutzki
Abstract: In this paper we study Secrecy-Preserving Query Answering problem under the Open World Assumption (OWA) for ELH Knowledge Bases. Here ELH is a top-free description logic ELH augmented with a modal operator ^. We employ a tableau procedure designed to compute a rooted labeled tree T which contains information about some assertional consequences of the given knowledge base. Given a secrecy set S, which is a finite set of assertions, we compute a function E, called an envelope of S, which assigns a set of assertions to each node of T. E provides logical protection to the secrecy set S against the reasoning of a querying agent. Once the tree T and an envelope E are computed, we define the secrecy-preserving tree TE. Based on the information available in TE, assertional queries with modal operator ^ can be answered efficiently while preserving secrecy. To the best of our knowledge, this work is first one studying secrecy-preserving reasoning in description logic augmented with modal operator ^. When the querying agent asks a query q, the reasoner answers “Yes” if information about q is available in TE; otherwise, the reasoner answers “Unknown”. Being able to answer “Unknown” plays a key role in protecting secrecy under OWA. Since we are not computing all the consequences of the knowledge base, answers to the queries based on just secrecy-preserving tree TE could be erroneous. To fix this problem, we further augment our algorithms by providing recursive query decomposition algorithm to make the query answering procedure foolproof.
Download
Paper Nr: 119
Title:

Strategy Composition in Dynamic Games with Simultaneous Moves
Authors:

Sujata Ghosh, Neethi Konar and R. Ramanujam
Abstract: Sometimes, in dynamic games, it is useful to reason not only about the existence of strategies for players, but also about what these strategies are, and how players select and construct them. We study dynamic games with simultaneous moves, repeated normal form games and show that this reasoning can be carried out by considering a single game, and studying composition of ``local'' strategies. We study a propostional modal logic in which such reasoning is carried out, and present complete axiomatization of the valid formulas.
Download
Paper Nr: 120
Title:

Agent-based Reconfigurable Natural Language Interface to Robots - Human-Agent Interaction using Task-specific Controlled Natural Languages
Authors:

Tamás Mészáros and Tadeusz Dobrowiecki
Abstract: We present the architecture of a flexible natural language based interface to robots involved in tasks in a mixed-initiative human-robot environment. The designed interface uses a bidirectional natural language communication pertinent to the user, the robot and the tasks at hand. Tasks are executed via agents that can communicate and engage in conversation with the user using task-specific controlled natural languages. The final interface language is dynamically composed and reconfigured at the interface level according to the tasks and skills of the robotic system. The multi-agent infrastructure is fused with the ROS robotic middleware to provide seamless communication to all system components at various level of abstraction.
Download
Paper Nr: 140
Title:

Optimized Non-visual Information for Deep Neural Network in Fighting Game
Authors:

Nguyen Duc Tang Tri, Vu Quang and Kokolo Ikeda
Abstract: Deep Learning has become most popular research topic because of its ability to learn from a huge amount of data. In recent research such as Atari 2600 games, they show that Deep Convolutional Neural Network (Deep CNN) can learn abstract information from pixel 2D data. After that, in VizDoom, we can also see the effect of pixel 3D data in learning to play games. But in all the cases above, the games are perfect-information games, and these images are available. For imperfect-information games, we do not have such bit-map and moreover, if we want to optimize our model by using only important features, then will Deep CNN still work? In this paper, we try to confirm that Deep CNN shows better performance than usual Neural Network (usual NN) in modeling Game Agent. By grouping important features, we increase the accuracy of modeling strong AI from 25.58% with a usual neural network to 54.24% with our best CNN structure.
Download
Paper Nr: 143
Title:

Deep Learning for Predictions in Emerging Currency Markets
Authors:

Svitlana Galeshchuk and Sumitra Mukherjee
Abstract: Accurate prediction of exchange rates is critical for devising robust monetary policies. Machine learning methods such as shallow neural networks have higher predictive accuracy than time series models when trained on input features carefully crafted by domain knowledge experts. This suggests that deep neural networks, with their ability to learn abstract features from raw data, may provide improved predictive accuracy with raw exchange rates as inputs. The preponderance of research focuses on developed currency markets. The paucity of research in emerging currency markets, and the crucial role that stable currencies play in such economies, motivates us to investigate the effectiveness of deep networks for exchange rate prediction in emerging markets. Literature suggests that the Efficient Market Hypothesis, which posits that asset prices reflect all relevant information, may not hold in such markets because of extraneous factors such as political instability and governmental interventions. This motivates our hypothesis that inclusion of carefully chosen macroeconomic factors as input features may improve the predictive accuracy of deep networks in emerging currency markets. This position paper proposes novel input features based on currency clusters and presents our method for investigating the hypothesis using exchange rates from developed as well as emerging currency markets.
Download
Paper Nr: 150
Title:

Clock-Model-Assisted Agent’s Spatial Navigation
Authors:

Joanna Isabelle Olszewska
Abstract: Intelligent agent’s navigation remotely controlled by means of natural language commands is of great help for robots operating in rescue activities or assistive aid. Whereas full conversation between the human commander and the agent could be limited in such situations, we propose thus to build human/robot dialogues based directly on semantically meaningful instructions like the directional spatial relations, in particular represented by the clock model, to efficiently communicate orders to the agent in the way it successfully gets to a target’s position. Experiments within real-world, simulated scenario have demonstrated the usefulness and effectiveness of our developed approach.
Download
Paper Nr: 3
Title:

Coordination, Synchronization and Localization Investigations in a Parallel Intelligent Robot Cellular Automata Model that Performs Foraging Task
Authors:

Danielli A. Lima, Claudiney R. Tinoco, Juan M. N. Viedman and Gina M. B. Oliveira
Abstract: Multiple agent systems can be applied to foraging tasks, thus solving this problem in a cooperative intelligent approach using cellular automata modeling. The objective is to construct an algorithm that performs foraging task correctly in Webots EDU simulation platform using robot architecture and also improves the individual controller model of each intelligent agent, using e-Puck devices properly. The proposed communication model has taken into account some cellular automata specifications, such as, the need for parallel synchronization, localization and accuracy of information dependency. After several simulations in Webots EDU, evaluating different approaches, the proposed communication model presented promising results on the parallel multi-robot foraging performance being pertinent in intelligent swarm robotics context.
Download
Paper Nr: 16
Title:

An Artificial Stock Market with Interactions Network and Mimetic Agents
Authors:

Sadek Benhammada, Frédéric Amblard and Salim Chikhi
Abstract: Agent-based artificial stock markets attracted much attention over the last years, and many models have been proposed. However, among them, few models take into account the social interactions and mimicking behaviour of traders, while the economic literature describes investors on financial markets as influenced by decisions of their peers and explains that this mimicking behaviour has a decisive impact on price dynamics and market stability. In this paper we propose a continuous double auction model of financial market, populated by heterogeneous traders who interact through a social network of influence. Traders use different investment strategies, namely: fundamentalists who make a decisions based on the fundamental value of assets; hybrids who are initially fundamentalists, but switch to a speculative strategy when they detect an uptrend in prices; noise traders who don’t have sufficient information to take rational decisions, and finally mimetic traders who imitate the decisions of their mentors on the interactions network. An experimental design is performed to show the feasibility and utility of the proposed model.
Download
Paper Nr: 19
Title:

Adjusting Word Embeddings by Deep Neural Networks
Authors:

Xiaoyang Gao and Ryutaro Ichise
Abstract: Continuous representations language models have gained popularity in many NLP tasks. To measure the similarity of two words, we have to calculate the cosine distances. However the qualities of word embeddings are due to the selected corpus. As for Word2Vec, we observe that the vectors are far apart to each other. Furthermore, synonym words with low occurrences or with multiple meanings are even further. In these cases, cosine similarities are not appropriate to evaluate how similar the words are. And considering about the structures of most of the language models, they are not as deep as we supposed. Based on these observations, we implement a mixed deep neural networks with two kinds of architectures. We show that adjustment can be done on word embeddings in both unsupervised and supervised ways. Remarkably, this approach improves the cases we mentioned by largely increasing almost all of synonyms similarities. It is also easy to train and adapt to certain tasks by changing the training target and dataset.
Download
Paper Nr: 35
Title:

The Octopus as a Model for Artificial Intelligence - A Multi-Agent Robotic Case Study
Authors:

Alfonso Íñiguez
Abstract: The aim of this paper is to investigate the curious cognition process exhibited by the octopus, and its practical applicability to multi-agent systems. The paper begins by explaining the limitations of using the human brain as a model to achieve artificial cognition and proposes an alternative model inspired by the octopus’ distributed approach to solving problems. As a case study, a laboratory prototype demonstrates awareness, autonomy, solidarity, expandability, and resiliency in a multi-robotic system. The cognition model described in this paper is primarily algorithmic and does not explore the model creation process nor semantics; rather, it lays the foundation and inspiration for a future realization as a Process for Agent Societies Specification and Implementation (PASSI).
Download
Paper Nr: 37
Title:

Development of an Intelligent Agent based Manufacturing System
Authors:

Hong-Seok Park and Ngoc-Hien Tran
Abstract: The new trend of the manufacturing system development is to apply autonomous behaviours inspired from biology for the manufacturing systems. In which, the resources of the manufacturing system are considered as biological organisms, which are autonomous entities so that the manufacturing system has the advanced characteristics inspired from biology such as self-adaptation, self-diagnosis, and self-optimization. To carry out these characteristics, the paper presents a paradigm about intelligent agent, called the cognitive agent and using cognitive agents for adapting to disturbances such as tool wear, machine breakdown that have happened on the shop floor. Modern manufacturing systems having the distributed control need autonomy and cooperation in solving problems of agents from agent technology, and cognitive capabilities for agents from cognitive technology. Cognitive agents combined from these two technologies are necessary for future manufacturing systems.
Download
Paper Nr: 49
Title:

Emoticon Recommendation System Reflecting User Individuality - A Preliminary Survey of Emoticon Use
Authors:

Taichi Matsui and Shohei Kato
Abstract: As the Internet has become widespread, text messaging has become a major means of communication. Because it is difficult to express emotion through text, emoticons were developed. There are many kinds of emoticons, and people often have difficulty finding one that conveys their meaning appropriately. This research aims to propose an emoticon recommendation system that considers individual differences. To this end, we conducted a survey about the use of emoticons. In this study, we report and analyze the results of this survey.
Download
Paper Nr: 54
Title:

Modified Krill Herd Optimization Algorithm using Focus Group Idea
Authors:

Mahdi Bidar, Edris Fattahi, Malek Mouhoub and Hamidreza Rashidy Kanan
Abstract: Krill Herd algorithm is one of most recently developed nature-inspired optimization algorithms which is inspired by herding behavior of krill individuals. In order to improve the performance of this algorithm to deal more effectively with high dimensional numerical functions, we propose a new method, called Focus Group idea to modify the solutions found by searching agents in group cooperation. In order to evaluate the effect of the proposed method on the performance of the Krill Herd algorithm, we conducted experiments on a set standard benchmark functions. The obtained results demonstrate the ability of the proposed method to improve the performance of the Krill Herd optimization algorithm.
Download
Paper Nr: 55
Title:

Ontology Learning Process as a Bottom-up Strategy for Building Domain-specific Ontology from Legal Texts
Authors:

Mirna El Ghosh, Hala Naja, Habib Abdulrab and Mohamad Khalil
Abstract: The objective of this paper is to present the role of Ontology Learning Process in supporting an ontology engineer for creating and maintaining ontologies from textual resources. The knowledge structures that interest us are legal domain-specific ontologies. We will use these ontologies to build legal domain ontology for a Lebanese legal knowledge based system. The domain application of this work is the Lebanese criminal system. Ontologies can be learnt from various sources, such as databases, structured and unstructured documents. Here, the focus is on the acquisition of ontologies from unstructured text, provided as input. In this work, the Ontology Learning Process represents a knowledge extraction phase using Natural Language Processing techniques. The resulted ontology is considered as inexpressive ontology. There is a need to reengineer it in order to build a complete, correct and more expressive domain-specific ontology.
Download
Paper Nr: 56
Title:

An Observation of Behavioral Changes of Indoor Dogs in Response to Caring Behavior by Humanoid Robots - Can Dogs and Robots Be Companions?
Authors:

Motoko Suzuki, Yuichi Sei, Yasuyuki Tahara and Akihiko Ohsuga
Abstract: The aim of our research is to build good relationships between pets and robots at home. We aim to promote of positive interaction between pets and robots. Recently, robots have been become popular with the general populace. There is a lot of research in human-robot interaction. We pay attention to pets that live in houses with humans. It is required for pets to like robots for positive interactions between pets and robots to exist. In this paper, we examine that 1) a robot can take care of dog, and 2) dogs and robots can be companion by caring behavior of robots toward dogs. In our experiment, we used two robots. One of the robots takes care of a dog, while the other does not. We observed which robot the dog chooses to interact with and had seventeen dogs participate in this study. We performed this statistical test to judge whether the dogs treated the robots with any significant differences.
Download
Paper Nr: 58
Title:

Searching the Optimal Combination of Fire Risks Reducing Measures at Oil and Gas Processing Facilities with the use of Genetic Algorithm
Authors:

Sergey Gudin, Renat Khabibulin and Denis Shikhalev
Abstract: The search for the combination of fire risk-reducing measures at oil and gas processing facilities is a complicated task. There may be a large number of measures to reduce fire risks which need to be optimized, both technically and economically. The analysis of the existing programs for risk assessment has been conducted. The structure of database with the values of risk-reducing measures has been worked out. To reduce the time required for this task, a genetic algorithm approach has been proposed.
Download
Paper Nr: 59
Title:

Practical Assumptions for Planning Under Uncertainty
Authors:

Juan Carlos Saborío and Joachim Hertzberg
Abstract: The (PO)MDP framework is a standard model in planning and decision-making under uncertainty, but the complexity of its methods makes it impractical for any reasonably large problem. In addition, task-planning demands solutions satisfying efficiency and quality criteria, often unachievable through optimizing methods. We propose an approach to planning that postpones optimality in favor of faster, satisficing behavior, supported by context-sensitive assumptions that allow an agent to reduce the dimensionality of its decision problems.We argue that a practical problem solving agent may sometimes assume full observability and determinism, based on generalizations, domain knowledge and an attentional filter obtained through a formal understanding of “relevance”, therefore exploiting the structure of problems and not just their representations.
Download
Paper Nr: 71
Title:

Towards Developing Dialogue Systems with Entertaining Conversations
Authors:

Hai-Long Trieu, Hiroyuki Iida, Nhien Pham Hoang Bao and Le-Minh Nguyen
Abstract: This paper explores a novel approach to developing a dialogue system that is able to make entertaining conversations with users. It proposes a method to improve the current goal-driven dialogue systems which support users for specific tasks while satisfying users’ goals with entertaining conversations. It then develops a dialogue system in which a set of features are considered to generate entertaining conversations, while reasonably prolonging the original too short dialogue. The game refinement measure is employed for the assessment of attractiveness since the conversations in dialogue systems can be seen as the process by which a player creates shoots or moves to win a game. The dialogues generated by the proposed method are evaluated by human subjects. The results confirm the effectiveness of the proposed method. The present idea can be a promising way to realize dialogue systems with entertaining conversations although further investigations are needed.
Download
Paper Nr: 73
Title:

Sarcasm Detection Method to Improve Review Analysis
Authors:

Shota Suzuki, Ryohei Orihara, Yuichi Sei, Yasuyuki Tahara and Akihiko Ohsuga
Abstract: Currently, classifying sarcastic sentences into positive and negative sentiments has been a difficult problem and an important task. The sarcastic sentences could indicate negative meaning by using positive expressions, or positive meaning by using negative expressions. Sarcasm is a special kind of sentiment that comprise of words which mean the opposite of what you really want to say, especially in order to insult or wit someone, to show irritation, or to be funny. Therefore, determining sarcasm is an important task in order to correctly classify the sentence. In this paper, we propose an approach to detect sarcasm. First, we apply dependency parsing to amazon review data. After that, we classify phrases in the sentence into the proposed phrase based on the sequence of part-of-speech as proposed by Bharti et al. After being classified into either one of the phrase types, it is determined whether each phrase is positive or negative. If the emotions of the situation phrases and the sentiment phrases are different, the sentence is determined to be a “sarcasm”. Using the above method, the experimental result shows the effectiveness of our method as compared with the the existing research.
Download
Paper Nr: 80
Title:

Evaluation of Local Descriptors for Automatic Image Annotation
Authors:

Ladislav Lenc
Abstract: Feature extraction is the first and often also the crucial step in many computer vision applications. In this paper we aim at evaluation of three local descriptors for the automatic image annotation (AIA) task. We utilize local binary patterns (LBP), patterns of oriented edge magnitudes (POEM) and local derivative patterns (LDP). These descriptors are successfully used in many other domains such as face recognition. However, the utilization of them in the AIA field is rather infrequent. The annotation algorithm is based on the K-nearest neighbours (KNN) classifier where labels from $K$ most similar images are ``transferred'' to the annotated one. We propose a label transfer method that assigns variable number of labels to each image. It is compared with an existing approach using constant number of labels. The proposed method is evaluated on three image datasets: Li photography, IAPR-TC12 and ESP. We show that the results of the utilized local descriptors are comparable to, and in many cases outperform the texture features usually used in AIA. We also show that the proposed label transfer method increases the overall system performance. The proposed method is evaluated on three image datasets: Li photography, IAPR-TC12 and ESP. We show that the results of the utilized local descriptors are comparable to, and in many cases outperform the texture features usually used in AIA. We also show that the proposed label transfer method increases the overall system performance.
Download
Paper Nr: 83
Title:

Data-driven Techniques for Expert Finding
Authors:

Veselka Boeva, Milena Angelova and Elena Tsiporkova
Abstract: In this work, we propose enhanced data-driven techniques that optimize expert representation and identify subject experts via automated analysis of the available online information. We use a weighting method to assess the levels of expertise of an expert to the domain-specific topics. An expert profile is presented by a description of the topics in which the person is an expert plus the relative levels (weights) of knowledge or experience he/she has in the different topics. In this context, we define a way to estimate the expertise similarity between experts. Then the experts finding task is viewed as a list completion task and techniques that return similar experts to ones provided by the user are considered. The proposed techniques are tested and evaluated on data extracted from PubMed repository.
Download
Paper Nr: 87
Title:

Data Mining and ANFIS Application to Particulate Matter Air Pollutant Prediction. A Comparative Study
Authors:

Mihaela Oprea, Marian Popescu, Sanda Florentina Mihalache and Elia Georgiana Dragomir
Abstract: The paper analyzes two artificial intelligence methods for particulate matter air pollutant prediction, namely data mining and adaptive neuro-fuzzy inference system (ANFIS). Both methods provide predictive knowledge under the form of rule base, the first method, data mining, as an explicit rule base, and ANFIS as an internal fuzzy rule base used to perform predictions. In order to determine the optimal number of prediction model inputs, we have perform a correlation analysis between particulate matter and other air pollutants. This operation imposed NO2 and CO concentrations as inputs of the prediction model, together with four values of PM10 concentration (from current hour to three hours ago), the output of the model being the prediction of the next hour PM10 concentration. The two prediction models are investigated through simulation in different structures and configurations using SAS® and MATLAB® respectively. The results are compared in terms of statistical parameters (RMSE, MAPE) and simulation time.
Download
Paper Nr: 109
Title:

Dynamic Agent-based Network Generation
Authors:

Audren Bouadjio-Boulic, Frederic Amblard and Benoit Gaudou
Abstract: Networks are a very convenient and tractable way to model and represent interactions among entities. For example, they are often used in agent-based models to describe agents’ acquaintances. Yet, data on real-world networks are missing or difficult to gather. Being able to generate synthetic but realistic social networks is thus an important challenge in social simulation. In this article, we provide a very comprehensive and modular agent-based process of network creation. We believe that the complexity of ABM (Agent-Based Models) comes from the overall interactions of entities, but they could be kept very simple for better control over the outcome. The idea is to use an agent-based simulation to generate networks: agent behaviors are rules for the network construction. Because we want the process to be dynamic and resilient to nodes perturbation, we provide a way for behaviors to spread among agents, following the meme basic principle - spreading by imitation. Resulting generated networks are compared to a target network; the system automatically looks at the best behavior distribution to generate this specific target network.
Download
Paper Nr: 115
Title:

Using Individual Feature Evaluation to Start Feature Subset Selection Methods for Classification
Authors:

Antonio Arauzo-Azofra, José Molina-Baena, Alfonso Jiménez-Vílchez and María Luque-Rodriguez
Abstract: Using a mechanism that can select the best features in a specific data set improves precision, efficiency and the adaptation capacity in a learning process and thus the resulting model as well. Normally, data sets contain more information than what is needed to generate a certain model. Due to this, many feature selection methods have been developed. Different evaluation functions and measures are applied and a selection of the best features is generated. This contribution proposes the use of individual feature evaluation methods as starting method for search based feature subset selection methods. An in-depth empirical study is carried out comparing traditional feature selection methods with the new started feature selection methods. The results show that the proposal is interesting as time gets reduced and classification accuracy gets improved.
Download
Paper Nr: 118
Title:

Valuing Others’ Opinions: Preference, Belief and Reliability Dynamics
Authors:

Sujata Ghosh and Katsuhiko Sano
Abstract: Deliberation often leads to changes in preferences and beliefs of an agent, influenced by the opinions of others, depending on how reliable these agents are according to the agent under consideration. Sometimes, it also leads to changes in the opposite direction, that is, reliability over agents gets updated depending on their preferences and/or beliefs. There are various formal studies of preference and belief change based on reliability and/or trust, but not the other way around $-$ this work contributes to the formal study of the latter aspect, that is, on reliability change based on agent preferences. In process, some policies of preference change based on agent reliabilities are also discussed. A two-dimensional hybrid language is proposed to describe such processes, and axiomatisations and decidability are discussed.
Download
Paper Nr: 121
Title:

Affinity-based Interpretation of Triangle Social Scenarios
Authors:

Pratyusha Kalluri and Pablo Gervás
Abstract: Computational interpretation of social scenarios is a critical step towards more human-like artificial intelligence. We present a model that interprets social scenarios by deducing the affinities of the constituent relationships. First, our model deploys Bayesian inference with an action affinity lexicon to infer probabilistic affinity relations characterizing the scenario. Subsequently, our model is able to use the inferred affinity relations to choose the most probable statement from multiple plausible statements about the scenario. We evaluate our approach on 80 Triangle-COPA multiple-choice problems that test interpretation of social scenarios. Our approach correctly answers the majority (59) of the 80 questions (73.75%), including questions about behaviors, emotions, social conventions, and complex constructs. Our model maintains interpretive power while using knowledge captured in the lightweight action affinity lexicon. Our model is a promising approach to interpretation of social scenarios, and we identify potential applications to automated narrative analysis, AI narrative generation, and assistive technology.
Download
Paper Nr: 122
Title:

An Analysis of Virtual Loss in Parallel MCTS
Authors:

S. Ali Mirsoleimani, Aske Plaat, Jaap van den Herik and Jos Vermaseren
Abstract: Monte Carlo tree search algorithms, such as UCT, select the best-root-child as a result of an iterative search process consistent with path dependency. Recent work has provided parallel methods that make the search process faster. However, these methods violate the path-dependent nature of the sequential UCT process. Here, a more rapid search thus results in a higher search overhead. The cost thereof is a lower time efficiency. The concept of virtual loss is proposed to compensate for this cost. In this paper, we study the role of virtual loss. Therefore, we conduct an empirical analysis of two methods for lock-free tree parallelization, viz. one without virtual loss and one with the virtual loss. We use the UCT algorithm in the High Energy Physics domain. In particular, we methodologically evaluate the performance of the both methods for a broad set of configurations regarding search overhead and time efficiency. The results show that using virtual loss for lock-free tree parallelization still degrades the performance of the algorithm. Contrary to what we aimed at.
Download
Paper Nr: 126
Title:

Graphics Processing Units for Constraint Satisfaction
Authors:

Malek Mouhoub and Ahmed Mobaraki
Abstract: A Constraint Satisfaction Problem (CSP) is a powerful formalism to represent constrained problems. A CSP includes a set of variables where each is defined over a set of possible values, and a set of relations restricting the values that the variables can simultaneously take. There are numerous problems that can be represented as CSPs. Solving CSPs is known to be quite challenging in general. The literature poses a great body of work geared towards finding efficient techniques to solve CSPs. These techniques are usually implemented in a system commonly referred to as a constraint solver. While many enhancements have been achieved over earlier ones, solvers still require powerful resources and techniques to solve a given problem in a reasonable running time. In this paper, a new parallel-based approach is proposed for solving CSPs. In particular, we design a new CSP solver that exploits the power of graphics processing units (GPU), which exist in modern day computers, as an affordable parallel computing architecture.
Download
Paper Nr: 127
Title:

Power Storage on the Smart Grid: Experimentation and Education
Authors:

Aaron Hunter and Ray Young
Abstract: If users can store electrical power, then it is possible to purchase power when it is cheap and use stored power when it is expensive. Spreading power consumption evenly in this manner can reduce costs and lower green house emissions. However, each user can define their own strategy and the interaction between these strategies can be complex. As a result, intelligent decision support tools can be useful in determining the most effective approach to power storage. In this paper, we describe the development of the Power Storage Simulator, a tool for experimenting with different storage algorithms. The simulator allows users to specify algorithms for charging and discharging stored power, and then observe the resulting changes in emissions and costs. We argue that there are at least two major benefits gained my allowing users to define storage strategies and experiment with simulated results. First, users can use the simulation to discover and validate effective storage algorithms. Second, the process of experimentation can actually educate consumers about the potential impact of effective power storage.
Paper Nr: 128
Title:

On the Replaceability of Computational Agents in an Ethical Theory
Authors:

Aaron Hunter
Abstract: In this paper, we are concerned with the use of existing ethical theories to analyze problems that involve both humans and computational agents. After introducing a handful of theories, we suggest that simply applying these theories to problems involving AI does not necessarily lead to novel insight. We argue, however, that existing ethical theories can be useful if we take a differential approach. In other words, for a fixed ethical theory, we try to identify cases where the presence of computational agents results in different ethical preferences. We suggest that a detailed analysis of such cases provides better understanding of the ethical issues at play, and it may even provide useful guidance in the design of ethical agents.
Paper Nr: 129
Title:

Deadlock Prevention in Rendezvous Generation for On-demand Inter-robot Resource Delivery
Authors:

Yin Chen, Xinjun Mao and Fu Hou
Abstract: In this paper, we consider a multi-robot system (MRS) which executes task points associated with 2-D locations. Each task point demands certain physical resources for its execution. A robot can fetch these resources either from fixed stations, or by conducting rendezvouses with other robots who happen to possess these resources, provided that the latter option can be more beneficial in terms of cost or resource availability. However, applying rendezvouses may cause deadlock among robots, through (1) the tangling among rendezvouses, or (2) sabotaging the resource consistency of the schedule of a robot which is originally holding the resources. We analyse these problems and introduce a series of deadlock prevention rules that are embedded into an A*-based rendezvous planning algorithm, so that both type of deadlocks can be avoided in the rendezvouses.
Download

Area 2 - Agents

Full Papers
Paper Nr: 4
Title:

Comparing Repair-Task-Allocation Strategies in MAS
Authors:

Hisashi Hayashi
Abstract: Many distributed systems can be regarded as multi-agent systems (MASs) where some agents are connected to a network but located in different places. We consider severe situations where many causes of future agent failures in MASs are found simultaneously and consecutively owing to large-scale disasters. If a cause of future agent failure is not removed within a limited time, there is a high possibility that one of the agents will stop working. In order to find effective strategies that reduce the number of actual agent failures, we compare some repair-task-allocation strategies for MASs where sensing agents find causes of future agent failures and manager agents communicate with one another to allocate repair tasks to action-execution agents.
Download
Paper Nr: 18
Title:

Utterance Behavior of Users While Playing Basketball with a Virtual Teammate
Authors:

Divesh Lala, Yuanchao Li and Tatsuya Kawahara
Abstract: Research on human-agent interaction has focused mainly on domains which are conversational in nature, but little work has been done on examining the behavior of interactive agents in domains such as team sports. This paper analyzes utterance behavior in this domain, specifically a virtual basketball game with an agent teammate. The main motivation is to assess the nature of utterances during the course of a game. We use a Wizard-of-Oz system which allows a hidden operator to appropriately respond to user utterances. Utterances are analyzed by annotating and categorizing according to Searle’s illocutionary speech acts. We find that there is evidence to support the process of the user beginning with basic utterances needed to play the game, confirming that the agent can understand them, and then moving to more complex utterances. We also find that non-task utterances are used and their proportion increases as the game progresses.
Download
Paper Nr: 38
Title:

Quantitative Robustness – A Generalised Approach to Compare the Impact of Disturbances in Self-organising Systems
Authors:

Jan Kantert, Sven Tomforde, Christian Müller-Schloer, Sarah Edenhofer and Bernard Sick
Abstract: Organic Computing (OC) and Autonomic Computing (AC) systems are distinct from conventional systems through their ability to self-adapt and to self-organise. However, these properties are just means and not the end. What really makes OC and AC systems useful is their ability to survive in a real world, i.e. to recover from disturbances and attacks from the outside world. This property is called robustness. In this paper, we propose a metric to gauge robustness in order to be able to quantitatively compare the effectiveness of different self-organising and self-adaptive system designs with each other. In the following, we apply this metric to three experimental application scenarios and discuss their usefulness.
Download
Paper Nr: 44
Title:

ε-Strong Privacy Preserving Multiagent Planner by Computational Tractability
Authors:

Jan Tožička, Antonín Komenda and Michal Štolba
Abstract: Classical planning can solve large and real-world problems, even when multiple entities, such as robots, trucks or companies, are concerned. But when the interested parties, such as cooperating companies, are interested in maintaining their privacy while planning, classical planning cannot be used. Although, privacy is one of the crucial aspects of multi-agent planning, studies of privacy are underepresented in the literature. A strong privacy property, necessary to leak no information at all, has not been achieved by any planner in general yet. In this contribution, we propose a multiagent planner which can get arbitrarily close to the general strong privacy preserving planner for the price of decreased planning efficiency. The strong privacy assurances are under computational tractability assumptions commonly used in secure computation research.
Download
Paper Nr: 50
Title:

Principles and Experiments of a Multi-Agent Approach for Large Co-Simulation Networks Initialization
Authors:

Jérémy Boes, Tom Jorquera and Guy Camilleri
Abstract: Simulating large systems, such as smart grids, often requires to build a network of specific simulators. Making heterogeneous simulators work together is a challenge in itself, but recent advances in the field of co-simulation are providing answers. However, one key problem arises, and has not been sufficiently addressed: the initialization of such networks. Many simulators need to have proper input values to start. But in the network, each input is another simulator’s output. One has to find the initial input values of all simulators such as their computed output is equal to the initial input value of the connected simulators. Given that simulators often contain differential equations, this is hard to solve even with a small number of simulators, and nearly impossible with a large number of them. In this paper, we present a mutli-agent system designed to solve the co-simulation initialization problem, and show preliminary results on large networks.
Download
Paper Nr: 60
Title:

Using Time Use Surveys in Multi Agent based Simulations of Human Activity
Authors:

Quentin Reynaud, Yvon Haradji, François Sempé and Nicolas Sabouret
Abstract: Human behavior simulations in multi agent systems often lack data to calibrate and qualify the representativeness of the simulated behaviors. In this paper, we will show that massive investigations such as time-use surveys allow us to obtain this type of data. At the present time, time-use surveys are mostly used to validate the realism of human activity at a macroscopic level (population scale). In this paper, we present a new method of human behavior generation that combines the use of time-use surveys to calibrate human activities, with a multi agent system enabling simulated behaviors to gain reactivity, autonomy, coordination and realism at a microscopic level (individual scale).
Download
Paper Nr: 84
Title:

Towards Collaborative Adaptive Autonomous Agents
Authors:

Mirgita Frasheri, Baran Cürüklü and Mikael Ekstroem
Abstract: Adaptive autonomy enables agents operating in an environment to change, or adapt, their autonomy levels by relying on tasks executed by others. Moreover, tasks could be delegated between agents, and as a result decision-making concerning them could also be delegated. In this work, adaptive autonomy is modeled through the willingness of agents to cooperate in order to complete abstract tasks, the latter with varying levels of dependencies between them. Furthermore, it is sustained that adaptive autonomy should be considered at an agent’s architectural level. Thus the aim of this paper is two-fold. Firstly, the initial concept of an agent architecture is proposed and discussed from an agent interaction perspective. Secondly, the relations between static values of willingness to help, dependencies between tasks and overall usefulness of the agents’ population are analysed. The results show that a unselfish population will complete more tasks than a selfish one for low dependency degrees. However, as the latter increases more tasks are dropped, and consequently the utility of the population degrades. Utility is measured by the number of tasks that the population completes during run-time. Finally, it is shown that agents are able to finish more tasks by dynamically changing their willingness to cooperate.
Download
Paper Nr: 113
Title:

A Study on Cooperative Action Selection Considering Unfairness in Decentralized Multiagent Reinforcement Learning
Authors:

Toshihiro Matsui and Hiroshi Matsuo
Abstract: Reinforcement learning has been studied for cooperative learning and optimization methods in multiagent systems. In several frameworks of multiagent reinforcement learning, the system’s whole problem is decomposed into local problems for agents. To choose an appropriate cooperative action, the agents perform an optimization method that can be performed in a distributed manner. While the conventional goal of the learning is the maximization of the total rewards among agents, in practical resource allocation problems, unfairness among agents is critical. In several recent studies of decentralized optimization methods, unfairness was considered a criterion. We address an action selection method based on leximin criteria, which reduces the unfairness among agents, in decentralized reinforcement learning. We experimentally evaluated the effects and influences of the proposed approach on classes of sensor network problems.
Download
Paper Nr: 134
Title:

Measuring Self-organisation at Runtime - A Quantification Method based on Divergence Measures
Authors:

Sven Tomforde, Jan Kantert and Bernard Sick
Abstract: The term “self-organisation” typically refers to the ability of large-scale systems consisting of numerous autonomous agents to establish and maintain their structure as a result of local interaction processes. The motivation to develop systems based on the principle of self-organisation is to counter complexity and to improve desired characteristics, such as robustness and context-adaptivity. In order to come up with a fair comparison between different possible solutions, a prerequisite is that the degree of self-organisation is quantifiable. Even though there are some attempts in literature that try to approach such a measure, there is none that is real-world applicable, covers the entire runtime process of a system, and considers agents as blackboxes (i.e. does not require internals about status or strategies). With this paper, we introduce a concept for such a metric that is based on external observations, neglects the internal behaviour and strategies of autonomous entities, and provides a continuous measure that allows for an easy comparibility.
Download
Paper Nr: 149
Title:

A Generic Agent Architecture for Cooperative Multi-agent Games
Authors:

João Marinheiro and Henrique Lopes Cardoso
Abstract: Traditional search techniques are difficult to apply to cooperative negotiation games, due to the often enormous search trees and the difficulty in calculating the value of a players position or move. We propose a generic agent architecture that ensembles negotiation, trust and opponent modeling, simplifying the development of agents capable of playing these games effectively by introducing modules to handle these challenges. We demonstrate the application of this modular architecture by instantiating it in two different games and testing the designed agents in a variety of scenarios; we also assess the role of the negotiation, trust and opponent modeling modules in each of the games. Results show that the architecture is generic enough to be applied in a wide variety of games. Furthermore, we conclude that the inclusion of the three modules allows for more effective agents to be built.
Download
Short Papers
Paper Nr: 13
Title:

An Agent-based Approach to Decentralized Global Optimization - Adapting COHDA to Coordinate Descent
Authors:

Joerg Bremer and Sebastian Lehnhoff
Abstract: Heuristics like evolution strategies have been successfully applied to optimization problems with rugged, multi-modal fitness landscapes, to non-linear problems, and to derivative free optimization. Parallelization for acceleration often involves domain specific knowledge for data domain partition or functional or algorithmic decomposition. We present an agent-based approach for a fully decentralized global optimization algorithm without specific decomposition needs. The approach extends the ideas of coordinate descent to a gossiping like decentralized agent approach with the advantage of escaping local optima by replacing the line search with a full 1-dimensional optimization and by asynchronously searching different parts of the search space using agents. We compare the new approach with the established covariance matrix adaption evolution strategy and demonstrate the competitiveness of the decentralized approach even compared to a centralized algorithm with full information access. The evaluation is done using a bunch of well-known benchmark functions.
Download
Paper Nr: 22
Title:

Multi-agent based Synchronous Communication for Dynamic Rescheduling in Railway Network
Authors:

Krishnendu Kundu and Animesh Dutta
Abstract: This paper presents a multi-agent based solution to change a predefined Railway schedule dynamically. This paper models a railway network consisting of passenger and freight trains, along with their fixed schedules following the rules of Indian Railways. The trains in the system are modelled as Train Agents and rail segments as Segment Agents. Delay of a single train due to unavoidable hazards like track related problems, rain, fog may cause severe delays for next trains in sequence. This paper proposes a method to suggest overtaking maneuvers for trains by passing different types of messages between Train Agents and Segment Agents. This paper proposes Selective Flooding of messages to avoid congestion overhead. All types of messages passed are elaborately described in the problem formulation section. This paper also contains a theoretical proof showing that Selective Flooding technique is always able to reach every node required. The proposed method has been simulated using JADE platform and the results are presented.
Download
Paper Nr: 53
Title:

Reactive Agent-based Model for Convergence of Autonomous Vehicles to Parallel Formations Heading to Predefined Directions of Motion
Authors:

Vander L. S. Freitas and Elbert E. N. Macau
Abstract: In this work we introduce a reactive agent-based model for convergence of autonomous vehicles to parallel formations heading to predefined directions of motion. They interact via rules of repulsion, alignment and attraction. There is also an abstraction of the desired path of motion, represented by a virtual guiding vehicle, which shows the desired direction to be followed by the formation. We performed simulations with different combinations of interaction rules and studied the parameter space. Additionally, we simulate the occurrence of communication failure among agents and the presence of noise. The resulting formations are evaluated by three quantifiers.
Download
Paper Nr: 66
Title:

On Interaction Quality in Human-Robot Interaction
Authors:

Suna Bensch, Aleksandar Jevtić and Thomas Hellström
Abstract: In many complex robotics systems, interaction takes place in all directions between human, robot, and environment. Performance of such a system depends on this interaction, and a proper evaluation of a system must build on a proper modeling of interaction, a relevant set of performance metrics, and a methodology to combine metrics into a single performance value. In this paper, existing models of human-robot interaction are adapted to fit complex scenarios with one or several humans and robots. The interaction and the evaluation process is formalized, and a general method to fuse performance values over time and for several performance metrics is presented. The resulting value, denoted interaction quality, adds a dimension to ordinary performance metrics by being explicit about the interplay between performance metrics, and thereby provides a formal framework to understand, model, and address complex aspects of evaluation of human-robot interaction.
Download
Paper Nr: 68
Title:

Applying the PROSA Reference Architecture to Enable the Interaction between the Worker and the Industrial Robot - Case Study: One Worker Interaction with a Dual-Arm Industrial Robot
Authors:

Ahmed R. Sadik and Bodo Urban
Abstract: Involving an industrial robot in a close physical interaction with the worker became quite possible, as a result of the availability of different collaborative industrial robots in the market. The physical cooperation between the industrial robot and the worker usually done under the umbrella of the flexible manufacturing paradigm, where both the industrial robot and the worker need to change their tasks fast and efficiently, to cope with the changes in the manufacturing process. This means that a reliable manufacturing control system must stand behind this physical interaction to achieve the proper communication interaction. A holonic control architecture is an ideal solution for this problem. Therefore, during this research we study the most commonly applied model of the holonic control architecture, then we apply this architecture on our case study, where one worker cooperates with a dual-arm industrial robot to build and produce any new product. Also the research uses the worker’s hand gesture recognition as a method to interact with the industrial robot during the execution of a cooperative production scenario.
Download
Paper Nr: 70
Title:

A Multi-agent Approach to Model and Analyze the Behavior of Vessels in the Maritime Domain
Authors:

Mathias Anneken, Yvonne Fischer and Jürgen Beyerer
Abstract: The automatic detection of suspicious behavior is one important part in order to support operators in surveillance tasks. Therefore, an approach to model the behavior of objects by using multi-agent systems is introduced. As each object has its own objectives and desires to fulfill, these are modeled as utility functions. The actions of the objects are estimated by using the Nash bargaining solution. Consequently, it is implied, that the objects are cooperating in order to achieve an optimal result for themselves. First results for this algorithm are shown by using examples from the maritime domain. On the one hand, the algorithm is used to calculate an anomaly score. On the other hand, it is used to predict the movement of vessels.
Download
Paper Nr: 79
Title:

Multi-Agent Approach for Evacuation Support System
Authors:

Shouhei Taga, Tomohumi Matsuzawa, Munehiro Takimoto and Yasushi Kambayashi
Abstract: We propose a system that supports evacuation after a large-scale disaster. When a large-scale disaster occurs, collecting information by using portable devices is difficult, because communication base stations collapse and traffic congestion occurs. Evacuees are at a loss where they should go for safe places in lack of information. In order to overcome these problems, we have proposed and evaluated a multi-agent system that is built on MANET. Our aim is to let the users share information on MANET, and provide safe routes to the destination by using collected information. In the previous paper, we proposed and implemented the main functions of the proposed system, and performed feasibility study of the multi-agent system by using simple simulator. In this paper, we report the revised system and investigate the relationships between the number of the users and diffusivity of information, and the number of the generated mobile agents by using newly constructed simulator. In the experiments for evaluation, we simulated a realistic situation by using real map data and we took stochastic change of the situation into consideration, because the situation must be getting worse by time elapses.
Download
Paper Nr: 103
Title:

Combining Behavioral Experiments and Agent-based Social Simulation to Support Trust-aware Decision-making in Supply Chains
Authors:

Diego de Siqueira Braga, Marco Niemann, Bernd Hellingrath and Fernando Buarque de L. Neto
Abstract: Trust is seen as one of the most important dimensions in developing and maintaining fruitful business relationships and has deep impact on the decision-making process in the supply chain planning. Despite its importance, very limited research has been done in the trust-aware decision-making field. This paper aims to experimentally examine how trust can be assessed over different dimensions and then be used to support decision-making in order to reduce the Bullwhip Effect, which is one of the biggest efficiency problems shown by supply chains of highly interconnected organizations. As industry is generally reluctant to provide data due to privacy concerns and trade secret protection, the authors of this paper, designed and conducted a web-based trust behavioral experiment. The data collected was used to evaluate the proposed trust mechanism through an Agent-Based Social Simulation. The results revealed that it is possible to infer trust relationships from behavioral experiments and historical based data, and use these relationships to influence the procurement, ordering and information sharing process. Although additional research is still necessary, the preliminary results revealed that the use of computational trust mechanisms can be helpful to lower the Bullwhip Effect.
Download
Paper Nr: 112
Title:

MP-ABT: A Minimal Perturbation Approach for Complex Local Problems
Authors:

Ghizlane El Khattabi, El Mehdi El Graoui, Imade Benelallam and El Houssine Bouyakhf
Abstract: The ability of Distributed Constraints Reasoning (DCR) to solve distributed combinatorial problems brings the DCR to have a considerable interest in multi-agent community. Hence, many DisCSP algorithms have been proposed in order to solve such distributed problems. The major limit of these algorithms is the simplification assumptions. The scientists assume that each agent is a simple one; it handles just one variable. But in the complex local problem case; where each agent has more than one variable; two methods are used: The compilation and the decomposition. These methods transform the original problem so as to make it as a simple one. In this paper, we propose a new protocol: MP-ABT (Minimal Perturbation complex local problems in the Asynchronous Backtracking). It is a resolution algorithm of DisCSPs with complex local problems. It is based on the ABT algorithm and the Dynamic CSP. Each complex agent is seen as a Minimal Perturbation Problem (MPP) and any received message is considered as a new intra-constraint perturbation event. The complex local problem is updated and a new MPP local solution is reported. The MP-ABT is presented and compared to three ABT families. Our experimental results show the MP-ABT effectiveness.
Download
Paper Nr: 114
Title:

Generating and Instantiating Abstract Workflows with QoS User Requirements
Authors:

Claudia Di Napoli, Luca Sabatucci, Massimo Cossentino and Silvia Rossi
Abstract: The growing availability of services accessible through the network makes it possible to build complex applications resulting from their composition that are usually characterized also by non-functional properties, known as Quality of Service (QoS). To exploit the full potential of service technology, automatic QoS-based composition of services is crucial. In this work a framework for automatic service composition is presented that relies on planning and service negotiation techniques for addressing both functional and non-functional requirements. The proposed approach allows for dynamic service composition and QoS attributes, and it can be applied when services are provided in the contest of a competitive market of service providers without knowledge disclosure.
Download
Paper Nr: 131
Title:

Development of Culture-specific Gaze Behaviours of Virtual Agents
Authors:

Tomoko Koda, Taku Hirano and Takuto Ishioh
Abstract: Gaze plays an important role in human-human communication. Adequate gaze control of a virtual agent is also essential for successful and believable human-agent interaction. Researchers in intelligent virtual agents have developed gaze control models by taking account of gaze duration, frequency and timing of gaze aversion. However, none of them have considered cultural differences in gaze behaviours. We aim to investigate cultural differences in gaze behaviours and their perception, by developing virtual agents with Japanese gaze behaviours, western gaze behaviours, their hybrid gaze behaviours, and full gaze behaviours, and compare their effects on the impressions of the agents and interactions. This position paper proposes our research agenda, describes the implemented gaze models, and our experimental design.
Download
Paper Nr: 135
Title:

Training Agents with Neural Networks in Systems with Imperfect Information
Authors:

Yulia Korukhova and Sergey Kuryshev
Abstract: The paper deals with multi-agent system that represents trading agents acting in the environment with imperfect information. Fictitious play algorithm, first proposed by Brown in 1951, is a popular theoretical model of training agents. However, it is not applicable to larger systems with imperfect information due to its computational complexity. In this paper we propose a modification of the algorithm. We use neural networks for fast approximate calculation of the best responses. An important feature of the algorithm is the absence of agent’s a priori knowledge about the system. Agents’ learning goes through trial and error with winning actions being reinforced and entered into the training set and losing actions being cut from the strategy. The proposed algorithm has been used in a small game with imperfect information. And the ability of the algorithm to remove iteratively dominated strategies of agents' behavior has been demonstrated.
Download
Paper Nr: 148
Title:

Emotion Contagion among Affective Agents - Issues and Discussion
Authors:

Mara Pudane, Michael A. Radin and Bernard Brooks
Abstract: Emotional contagion is a mechanism which results in transferring an emotion from one person to another; in fact, it has been proven to be one of the key factors in successful crowd managing and preserving an amiable working atmosphere. However, believable simulation of a group of people with social links amongst them requires not only complex interaction models but also complex internal models as well. This paper describes the progress of an on-going research that explores and simulates various types of emotions while they are being transmitted amongst affective intelligent agents that are connected in simulated social structure. The internal mechanism of agents is based on affective agent architectures while the contagion and its rules are being modelled by using tools from graph theory.
Download
Paper Nr: 154
Title:

A Global Path Planning Strategy for a UGV from Aerial Elevation Maps for Disaster Response
Authors:

D. C. Guastella, L. Cantelli, C. D. Melita and G. Muscato
Abstract: An approach for global path planning of an Unmanned Ground Vehicle (UGV) is proposed, including basic traversability analysis of the rough terrain to get through. The navigation capabilities of the UGV, in performing such analysis, are considered. The here proposed solution is organized into two following phases: first an aerial scan of the environment is executed by a UAV (Unmanned Aerial Vehicle) and the elevation map of the area is built; after that, a set of processing algorithms is applied to such surface model to derive a 2D costmap (whose costs are based on the prior traversability analysis) which is given as input of a D* path planner. The resulting path can be eventually delivered as a sequence of waypoints for a navigation controller on the field mobile platform.
Download
Paper Nr: 156
Title:

Simulation of Language Evolution based on Actual Diachronic Change Extracted from Legal Terminology
Authors:

Makoto Nakamura, Yuya Hayashi and Ryuichi Matoba
Abstract: Simulation studies have played an important role in language evolution. Although a variety of methodologies have been proposed so far, they are typically too abstract to recognize that their learning mechanisms properly reflect actual ones. One reason comes from the lack of empirical data recorded for a long period with explicit description. Our purpose in this paper is to show simulation models adapt to actual language change. As empirical diachronic data, we focus on a statutory corpus. In general, statutes define important legal terms with explanatory sentences, which are also revised by amendment. We proposed an iterated learning model, in which an infant agent learns grammar through his/her parent’s utterances about legal terms and their semantic relations, and the infant becomes a parent in the next generation. The key issue is that the learning situation about legal terms and their relations can be changed due to amendment. Our experimental result showed that infant agents succeeded to acquire compositional grammar despite irregular changes in their learning situation.
Download
Paper Nr: 5
Title:

Autonomous Agents in Multiagent Organizations
Authors:

Marc Premm and Stefan Kirn
Abstract: Autonomous agents are constantly gaining relevance in economic applications. Autonomy as a characteristic of agents enables flexible behavior in cases of unforeseeable conditions. Multiagent systems research has analyzed various dimensions of autonomous behavior. However, the application of agents in an organizational context requires the actors to apply externally given rules that restrict agent autonomy. While multiagent systems aim at maximum flexibility, economical applications in organizations require stable structures. Multiagent organizations in terms of structured and stable multiagent systems are necessary to successfully link autonomous agents with organizations. Modelling autonomous agents in multiagent organizations requires to include the organizational structure and the operational processes, but also needs to consider the constitutive processes that enable the creation, adaption and dissolution of multiagent organizations. We survey extant literature from distributed artificial intelligence and management science and propose models for organizational structure and procedure of multiagent organizations. The models address new aspects for including autonomous agents in organizations that result from the linkage between both perspectives.
Download
Paper Nr: 27
Title:

Generating Swarm Solution Classes using the Hamiltonian Method of Swarm Design
Authors:

M. Li, C. Qiu, J. Park, D. Chan, J. Jeon, J. Na, C. Wong, B. Zhao, E. Chang, S. Kazadi and S. Hettiarachchi
Abstract: We utilize a swarm design methodology that enables us to develop classes of swarm solutions to specific specifications. The method utilizes metrics devised to evaluate the swarm’s progress – the global variables – along with the set of available technologies in order to answer varied questions surrounding a swarm design for the task. These questions include the question of whether or not a swarm is necessary for a given task. The Jacobian matrix, here identified as the technology matrix, is created from the global variables. This matrix may be interpreted in a way that allows the identification of classes of technologies required to complete the task. This approach allows us to create a class of solutions that are all suitable for accomplishing the task. We demonstrate this capability for accumulation swarms, generating several configurations that can be applied to complete the task. If the technology required to complete the task either cannot be implemented on a single agent or is unavailable, it may be possible to utilize a swarm to generate the capability in a distributed way. We demonstrate this using a gradient-based search task in which a minimal swarm is designed along with two additional swarms, all of which extend the agents’ capabilities and successfully accomplish the task.
Download
Paper Nr: 51
Title:

DM-UAV: Dexterous Manipulation Unmanned Aerial Vehicle
Authors:

Alberto Torres, Francisco Candelas, Damian Mira and Fernando Torres
Abstract: This paper describes a novel aerial manipulation system, a DM-UAV which is composed of a drone with a robotic hand. The main objective is to grasp a target object with the robotic hand. We assume that the object position is known so the drone flies to this position and lands, then the robotic hand can grasp the target object. After of that, the drone can take off in order to transport the object to other location. This system can be very useful for different field applications, f.e. agriculture, to clean the trash on the field, to eliminate contaminating objects, etc.
Download
Paper Nr: 52
Title:

Multi-agent Polygon Formation using Reinforcement Learning
Authors:

B. K. Swathi Prasad, Aditya G. Manjunath and Hariharan Ramasangu
Abstract: This work provides details a simulation experiment and analysis of Q-learning applied to multi-agent systems. Six agents interact within the environment to form hexagon, square and triangle, by reaching their specific goal states. In the proposed approach, the agents form a hexagon and the maximum dimension of this pattern is be reduced to form patterns with smaller dimensions. A decentralised approach of controlling the agents via Q-Learning was adopted which reduced complexity. The agents will be able to either move forward, backward and sideways based on the decision taken. Finally, the Q-Learning action-reward system was designed such that the agents could exploit the system which meant that they would earn high rewards for correct actions and negative rewards so the opposite.
Download
Paper Nr: 62
Title:

Cooperative Multi-agent Approach for Computational Systems of Systems Architecting
Authors:

Teddy Bouziat, Stéphanie Combettes, Valérie Camps and Jeremy Boes
Abstract: This paper addresses the modeling and design of Systems of Systems (SoS). It presents and illustrates a new generic model to describe formally such systems. This model is used to propose a SoS architecting approach based on adaptive multi-agent systems. In this approach, each component system composing the SoS uses a local cooperative decision process in order to interact with other systems and to collectively give rise to a relevant overall function at the SoS level. The proposed model as well as the proposed approach are instantiated with a simulated unmanned aerial vehicle scenario and compared with another approach dealing with collaboration between systems in a SoS.
Download
Paper Nr: 75
Title:

Comparison of Agents’ Performance in Learning to Cross a Highway for Two Decisions Formulas
Authors:

Anna T. Lawniczak and Fei Yu
Abstract: We compare the performance of simple cognitive agents, learning to cross a Cellular Automaton (CA) based highway, for two decision formulas used by the agents’ in their decision-making process. We describe the main features of the simulation model: CA based highway traffic environment, agents and their decision and learning mechanisms. The agents use a type of “observational social learning” strategy, i.e. they observe the performance of other agents and they try to mimic what worked for other agents and they try to avoid what did not work for the other agents. In the decision-making process of deciding whether to cross the highway or to wait, depending on the simulation setup, the agents use one of the two decisions formulas: the first one based only on the assessment of the agents crossing decisions (cDF), or the second one based on the assessment of the agents crossing and waiting decisions (cwDF). Our simulations show that the performance of agents using cwDF is much better than the performance of the agents using cDF in their decision making process. We measure the agents’ performance by the numbers of agents: who crossed successfully, who were killed and those who are still queuing to cross at simulation end.
Download
Paper Nr: 86
Title:

Fault Tolerance Analysis for Dependable Autonomous Agents using Colored Time Petri Nets
Authors:

Lan Anh Trinh, Baran Cürüklü and Mikael Ekström
Abstract: Fault tolerance has become more and more important in the development of autonomous systems with the aim to help the system to recover its normal activities even when some failures happen. Yet, one of the concerns is how to analyze the reliability of a fault tolerance mechanism with regards to the collaboration of multiple agents to complete a complicated task. To do so, an approach of fault tolerance analysis with the colored time Petri net framework is proposed in this work, where a task can be represented by a tree of different concurrent and dependent subtasks to assign to agents. Different subtasks and agents are modeled by color tokens in Petri network. The time values are added to evaluate the processing performance of the whole system with respect to its ability to solve a task with fault tolerance ability. The colored time Petri nets are then tested with simulation of centralized and distributed systems. Finally the experiments are performed to show the feasibility of the proposed approach. From the basics of this study, a generalized framework in the future can be developed to address the fault tolerance analysis for a set of agents working with a sophisticated plan to achieve a common target.
Download
Paper Nr: 93
Title:

An Agent-based Simulation of Extremist Network Formation through Radical Behavior Diffusion
Authors:

Carlos Sureda, Benoit Gaudou and Frederic Amblard
Abstract: Understanding how terrorist networks are created and how individuals turn into extremism and then into terrorism is a current subject of interest and a cross-domain research problem since it involves scholars from political sciences, sociology, physics and computer scientists among others. In this paper, an agent-based approach is used to simulate the process of radicalization and creation of a terrorist network, and the link between both processes. Each citizen has several attributes allowing the model to take into account heterogeneous profiles of individual. Furthermore, we model the social transfer that takes place during the interaction of individuals in order to understand how cultural ideas are transmitted. This paper also provides a non-exhaustive but detailed survey of the state of the art on the agent-based terrorist networks modelling.
Download
Paper Nr: 98
Title:

Efficient Exploration of Environments Using Stochastic Local Search
Authors:

Ramoni O. Lasisi
Abstract: This research provides a preliminary investigation of the use of Stochastic Local Search (SLS) technique to explore complex environments where agents' knowledge and the time to explore such environments are limited.~We model the problem as that of an instance of a search problem and develop SLS technique that enables efficient exploration of such relatively difficult environments by teams of agents.~Results from experiments using teams of various sizes that implement the proposed model show the effectiveness of the technique.~In most cases of the problem instances, teams of agents were able to complete exploration of more than 70% of the environments.~While in the best cases, they were able to complete explorations of more than 80% of the environments within short period of time.
Paper Nr: 102
Title:

An Off-line Evaluation of Users’ Ranking Metrics in Group Recommendation
Authors:

Silvia Rossi, Francesco Cervone and Francesco Barile
Abstract: One of the major issue in designing group recommendation techniques relates to the difficulty of the evaluation process. Up-today, no freely available dataset exists that contains information about groups, like, for example, the group’s choices or social aspects that may characterize the group’s members. The objective of the paper is to analyze the possibility to make an evaluation of ranking-based groups recommendation techniques by using offline testing. Typically, the evaluation of group recommendations is computed, as in the classical single user case, by comparing the predicted group’s ratings with respect to the single users’ ratings. Since the information contained in the datasets are mainly such user’s ratings, here, ratings are used to define different ranking metrics. Results suggest that such an attempt is hardly feasible. Performance seems not to be affected by the choice of ranking technique, except for some particular cases. This could be due to the averaging effect of the evaluation with respect to the single users’ ratings, so a deeper analysis or specific dataset are necessary.
Download
Paper Nr: 116
Title:

Nested Rollout Policy Adaptation for Multiagent System Optimization in Manufacturing
Authors:

Stefan Edelkamp and Christoph Greulich
Abstract: In manufacturing there are not only flow lines with stations arranged one behind the other, but also more complex networks of stations where assembly operations are performed. The considerable difference from sequential flow lines is that a partially ordered set of required components are brought together in order to form a single unit at the assembly stations in a competitive multiagent system scenario. In this paper we optimize multiagent control for such flow production units with recent advances of Nested Monte-Carlo Search. The optimization problem is implemented as a single-agent game in a generic search framework. In particular, we employ Nested Monte-Carlo Search with Rollout Policy Adaptation and apply it to a modern flow production unit, comparing it to solutions obtained with a simulator and with a model checker.
Download
Paper Nr: 136
Title:

Toward Gamified Knowledge Contents Refinement - Case Study of a Conversation Partner Agent
Authors:

Takayuki Iwamae, Kazuhiro Kuwabara and Hung-Hsuan Huang
Abstract: Rich knowledge contents are necessary to develop an intelligent agent that interacts with people and supports their communication or their activities. In this paper, we choose a conversation partner agent for people with aphasia as an example and propose a method that interactively acquires and refines knowledge contents for the agent. The proposed method is invoked when a problem is found in the knowledge contents and utilizes the concept of gamified crowdsourcing. Gamified tasks verify the data input by a user. By utilizing a crowdsourcing approach, we strive for more accurate knowledge contents. The paper presents its game design and an example scenario.
Download
Paper Nr: 137
Title:

A 3D Anti-collision System based on Artificial Potential Field Method for a Mobile Robot
Authors:

Carlos Morais, Tiago Nascimento, Alisson Brito and Gabriel Basso
Abstract: Anti-collision systems are based on sensing and estimating the mobile robot pose (coordinates and orientation), with respect to its environment. Obstacles detection, path planning and pose estimation are primordial to ensure the autonomy and safety of the robot, in order to reduce the risk of collision with objects and living beings that share the same space. For this, the use of RGB-D sensors, such as the Microsoft Kinect, has become popular in the past years, for being relative accurate and low cost sensors. In this work we proposed a new 3D anti-collision algorithm based on Artificial Potential Field method, that is able to make a mobile robot pass between closely spaced obstacles, while also minimizing the oscillations during the cross. We develop our Unmanned Ground Vehicles (UGV) system on a ’Turtlebot 2’ platform, with which we perform the experiments.
Download
Paper Nr: 139
Title:

An Electricity Market Game using Agent-based Gaming Technique for Understanding Energy Transition
Authors:

Setsuya Kurahashi and Wander Jager
Abstract: The Electricity Market in Japan has been an oligopolistic market since the previous century, but it will be a liberalised competitive market soon due to a policy change. It is supposed to provide wholesale power markets. Therefore, it has high possibilities to become two-sided markets with strong wholesalers. The goal of this study is to clarify decisive factors for making decision of energy selection based on human competitive and collaboration behaviour to be helpful for an incentive design of energy markets. For the purpose, two hypotheses were set in the experiment. First is that energy transition to renewable source is achieved by players while keeping their profit. Second is that aggregators have ability to control the energy market through the share of consumers' power market as well as other two-sided markets. Our experiment confirmed that the energy orientation of electric power consumers could give a significant influence on power generation investment of electric power suppliers, and the risk of nuclear energy was underestimated. And the first hypothesis was adopted and the second was rejected by the experiments through the agent-based gaming.
Download
Paper Nr: 144
Title:

An App-based Algorithmic Approach for Harvesting Local and Renewable Energy using Electric Vehicles
Authors:

Antoine Dubois, Antoine Wehenkel, Raphael Fonteneau, Frédéric Olivier and Damien Ernst
Abstract: The emergence of electric vehicles (EVs), combined with the rise of renewable energy production capacities, will strongly impact the way electricity is produced, distributed and consumed in the very near future. This position paper focuses on the problem of optimizing charging strategies for a fleet of EVs in the context where a significant amount of electricity is generated by (distributed) renewable energy. It exposes how a mobile application may offer an efficient solution for addressing this problem. This app can play two main roles. Firstly, it would incite and help people to play a more active role in the energy sector by allowing photovoltaic (PV) panel owners to sell their electrical production directly to consumers, here the EVs’ agents. Secondly, it would help distribution system operators (DSOs) or transmission system operators (TSOs) to modulate more efficiently the load by allowing them to influence EV charging behaviour in real time. Finally, the present paper advocates for the introduction of a two-sided market-type model between EV drivers and electricity producers.
Download