Multi-Agent Systems and Agreement Technologies : 17th European Conference, EUMAS 2020, and 7th International Conference, AT 2020, Thessaloniki, Greece, September 14-15, 2020, Revised Selected Papers

This book constitutes the revised post-conference proceedings of the 17th European Conference on Multi-Agent Systems, EUMAS 2020, and the 7 th International Conference on Agreement Technologies, AT 2020, which were originally planned to be held as a joint event in Thessaloniki, Greece, in April 2020. Due to COVID-19 pandemic the conference was postponed to September 2020 and finally became a fully virtual conference. The 38 full papers presented in this volume were carefully reviewed and selected from a total of 53 submissions. The papers report on both early and mature research and cover a wide range of topics in the field of autonomous agents and multi-agent systems. Preface Organization Contents EUMAS 2020 Session 1: Intelligent Agents and MAS Applications Towards a Theory of Intentions for Human-Robot Collaboration 1 Introduction 2 Related Work 3 Cognitive Architecture 3.1 Action Language and Domain Representation 3.2 Adapted Theory of Intention 3.3 Refinement, Zooming and Execution 4 Experimental Setup and Results 4.1 Experimental Results (Simulation) 4.2 Execution Trace 4.3 Robot Experiments 5 Discussion and Future Work References Decentralised Control of Intelligent Devices: A Healthcare Facility Study 1 Introduction 2 Related Work 3 Problem Statement 4 DCOPs for Device Management 4.1 Room-Based Approach 4.2 Area-Based Approach 4.3 Multi-variable Area-Based Approach 5 A DPOP Solution for the Device Management Problem 6 Empirical Evaluation 6.1 Benchmarking 6.2 Message Size 7 Conclusion References Decentralised Multi-intersection Congestion Control for Connected Autonomous Vehicles 1 Introduction 2 Background on Intersection Model and Rules for Vehicles 3 A Space-Efficient Intersection Model 3.1 Structural Constraints 3.2 Objective of Each Intersection and Discussion 4 Priority Levels for Multi-intersection Settings 4.1 Calculating Priority Levels 5 DCOPs for Intersection Management 5.1 Optimisation 5.2 The Max-Sum_AD_VP Algorithm and the Importance of Node Ordering 6 Empirical Evaluation 6.1 Evaluating Space Efficiency at Individual Intersections 6.2 Evaluating the Efficiency of the Max-Sum_AD_VP Algorithm at Individual Intersections 6.3 Multi-intersection Efficiency 6.4 Discussion on the Lane-Based Approach 7 Conclusions References Congestion Management for Mobility-on-Demand Schemes that Use Electric Vehicles 1 Introduction 2 Related Work 3 Problem Definition 4 Optimal Offline Scheduling 5 Greedy Online Scheduling 6 Evaluation 6.1 EXP1: Customer Service 6.2 EXP2: Execution Time and Scalability 7 Conclusions and Future Work References Disaster Response Simulation as a Testbed for Multi-Agent Systems 1 Introduction 2 Related Work 3 A Simulator for Disaster Response Episodes 3.1 Problem Set-Up 3.2 Architecture of the Disaster Response Simulator 3.3 Communication with Multi-Agent Platforms 3.4 Simulation Cycle 3.5 Simulation Metrics 4 Evaluating the Simulator 4.1 Simulator Performance 4.2 Comparison Between Different MAS Approaches 5 Conclusions and Future Work References EUMAS 2020 Session 2: Mechanisms, Incentives, Norms, Privacy Rewarding Miners: Bankruptcy Situations and Pooling Strategies 1 Introduction 1.1 Mining Pool Attacks 1.2 Related Works on Rewarding Mechanisms 2 Preliminaries 2.1 The Model 2.2 Bankruptcy Situations 3 Incentive Compatible Reward Functions 4 A Multi-pool Analysis 4.1 Hopping Analysis on Schrijver's Rewarding Function 4.2 Hopping Analysis on CEL-Based Rewarding Function 4.3 Comparison of the Two Rewarding Functions in a Multi-pool Framework 5 Conclusion References A Game-Theoretical Analysis of Charging Strategies for Competing Double Auction Marketplaces 1 Introduction 2 Related Works 3 Framework 3.1 Basic Setting 3.2 Trader's Expected Utility 3.3 The Marketplace's Expected Utility 4 Solving the Nash Equilibrium Charging Strategy 4.1 The Fictitious Play Algorithm 5 Nash Equilibrium Strategies of Sellers and Buyers 6 Equilibrium Analysis of Marketplace's Charging Strategies 7 Conclusion References Agents for Preserving Privacy: Learning and Decision Making Collaboratively 1 Introduction 2 Agent-Based Auctioning for Privacy: PANO 2.1 Background: Clarke-Tax Mechanism 2.2 PANOAuctions 2.3 Privacy Policy 3 Learning to Bid 3.1 Bidding Ranges 3.2 Effective Auctions 3.3 Utility Update 4 Evaluation of Learning for Preserving Privacy 4.1 Simulation System 4.2 PANOLA vs. PANO Agents 4.3 Exploration Within Bid Ranges 5 Discussion References Open Social Systems 1 Introduction 2 Proposal 2.1 Norms 2.2 Profiles 3 Notation 4 Architecture and Associated Operational Model 5 Motivating Example 6 Conclusion References A Faithful Mechanism for Privacy-Sensitive Distributed Constraint Satisfaction Problems 1 Introduction 2 Related Literature 3 Multi-agent Meeting Scheduling 4 Privacy Leakage 5 A Negotiation-Based Incentive Mechanism 6 Properties of the Mechanism 6.1 Responders' Faithfulness 6.2 Initiator's Faithfulness 7 Conclusions References Incentivising Exploration and Recommendations for Contextual Bandits with Payments 1 Introduction 2 Problem Statement 3 Algorithms and Guarantees 3.1 Other Payments Scheme and Lower Bound 4 Simulations 5 Conclusion References Emotional Agents Make a (Bank) Run 1 Introduction 2 Background and Related Work 2.1 Emotions 2.2 Emotional Agents in Socioeconomic Scenarios 2.3 Bank Runs 3 Modelling Agents Using X-Machines and Emotions 3.1 A Formal Model of Agents 3.2 A Formal Model of Emotions 3.3 A Formal Model of Emotional Agents 4 Modelling Bank Runs 4.1 Environment Setup 4.2 Agent Parameters Setup 5 Experimental Results 5.1 Calibration 5.2 The Effect of Influencers 6 Conclusions References EUMAS 2020 Session 3: Autonomous Agents An Interface for Programming Verifiable Autonomous Agents in ROS 1 Introduction 2 Background 3 Integrating Autonomous Agents with ROS 3.1 Connecting to Rosbridge 3.2 Subscribing 3.3 Publishing 4 Case Studies 4.1 TurtleBot Autonomous Patrolling 4.2 The Three TurtleBots: Home Service Robots 5 Related Work 6 Conclusion References Integrated Commonsense Reasoning and Deep Learning for Transparent Decision Making in Robotics 1 Introduction 2 Related Work 3 Architecture 3.1 Knowledge Representation, Reasoning, and Learning 3.2 Relational Descriptions as Explanations 4 Experimental Setup and Results 4.1 Experimental Setup 4.2 Execution Trace 4.3 Experimental Results 5 Conclusions References Combining Lévy Walks and Flocking for Cooperative Surveillance Using Aerial Swarms 1 Introduction 1.1 Contribution 2 Swarm Systems 2.1 Surveillance in Swarm Systems 2.2 Lévy Walks in Swarm Systems 3 Proposed Model 3.1 Flocking:Interaction 3.2 Lévy Walk 4 Experiments and Results 4.1 Simulation Experiments 4.2 Preliminary Real Experiments 5 Conclusions and Future Work References Single-Agent Policies for the Multi-Agent Persistent Surveillance Problem via Artificial Heterogeneity 1 Introduction and Background 2 Persistent Surveillance Problem 2.1 PSP Simulation Environment 2.2 Local Observations 2.3 Action Policy 2.4 Reward Function 2.5 Analytical Assessment 3 Single-Agent Policies for the PSP 3.1 Five Policies for PSP 3.2 Multi-Agent Deployment of Single-Agent Policies 4 Results and Discussion 4.1 Homogeneous-Policy Performance 4.2 Homogeneous-Policy Convergence Cycle 4.3 Artificial Heterogeneity 5 Conclusion References Explaining the Influence of Prior Knowledge on POMCP Policies 1 Introduction 2 Related Work 3 Materials and Methods 3.1 Partially Observable Markov Decision Processes 3.2 POMCP 3.3 Problem Formalization 3.4 Planning Strategies 3.5 Experimental Setup 3.6 Measures for Policy Explanation 4 Results 5 Conclusion and Ongoing Work References EUMAS 2020 Best Papers Session Approximating Voting Rules from Truncated Ballots 1 Introduction 2 Preliminaries 3 Approximating Voting Rules from Truncated Ballots 3.1 Borda and Positional Scoring Rules 3.2 Rules Based on Pairwise Comparisons 4 Probability of Selecting the True Winner 4.1 Experiments Using Mallows Model 4.2 Experiments Using Real Data Sets 5 Measuring the Approximation Ratio 5.1 Worst Case Study 5.2 Average Case Evaluation 5.3 Real Data Sets 6 Conclusion References Privacy-Preserving Dialogues Between Agents: A Contract-Based Incentive Mechanism for Distributed Meeting Scheduling 1 Introduction 2 Meeting Scheduling Problem Formulation 2.1 Problem Definition 2.2 Agents' Preferences and Privacy Leakage 3 Contract-Based Meeting Scheduling Protocol 4 Two-Layer Contract Design 4.1 Joint Reward and Privacy Leakage Optimal Control 4.2 Properties of the Mechanism 5 Simulation Results 6 Conclusions A Appendix: Proof of Proposition 1 B Appendix: Proof of Proposition 2 References EUMAS-AT 2020 Joint Session An Argumentation-Based Approach to Generate Domain-Specific Explanations 1 Introduction 2 Related Work 3 A Formal Model to Represent Argumentation Frameworks 3.1 Formal Model 3.2 Mapping ASPIC+ Theory into Our Formal Model 3.3 The EvalAF Algorithm 3.4 The ExpAF Algorithm 4 Arguments and Attacks 4.1 Guideline Representation 4.2 Patient Information 4.3 Argument Schemes 4.4 Attack Schemes 5 A Stroke Survivor: Baula 5.1 Explanations 5.2 Demonstration of the Proposed Approach 6 Discussion and Conclusion References Distributed Multi-issue Multi-lateral Negotiation Using a Divide and Rule Approach 1 Introduction 2 Negotiation Framework 3 Negotiation Protocol 3.1 Clustering Phase 3.2 Negotiation Phase 4 Negotiation Tactics 4.1 Methods to Compute ei 4.2 Methods to Compute a Range of Values for an Attribute 4.3 Common Tactic Defined by an Alliance 4.4 Negotiation Outcome 5 Theoretical Analyse 5.1 Negotiation in the Best Case 5.2 Negotiation in the Worst Case 6 Experimental Results 7 Conclusion References Increasing Negotiation Performance at the Edge of the Network 1 Introduction 2 Context and Motivating Examples 3 Background in Alternating Offers Protocol 4 Our Proposal: Alternating Constrained Offers Protocol 5 Experimental Methodology 5.1 Problem Generation 5.2 Running the Simulations 6 Results 6.1 Impact of Adopting ACOP on Negotiation Length 6.2 Impact of Adopting ACOP on Competitive Advantage 7 Conclusion References Challenges and Main Results of the Automated Negotiating Agents Competition (ANAC) 2019 1 Introduction 2 Automated Negotiation Main League 3 Human-Agent League 4 The Diplomacy League 5 The Werewolf League 6 Supply Chain Management 7 Future Directions References Optimal Majority Rule Versus Simple Majority Rule 1 Introduction 1.1 The Model 1.2 Related Work and Contribution 2 Optimal Acceptance Threshold 2.1 A General Result 2.2 Proposals Generated by Continuous Uniform Distributions 2.3 Proposals Generated by Normal Distributions 2.4 Proposals Generated by Symmetrized Pareto Distributions 2.5 Proposals Generated by Laplace Distributions 2.6 Proposals Generated by Logistic Distributions 3 Comparison of the Expected Utility Increments 4 Conclusion References Evaluating Crowdshipping Systems with Agent-Based Simulation 1 Introduction 2 Related Work 3 Crowdshipping Model 4 Crowdshipping Simulator 4.1 Simulation Engine 4.2 Simulator Configuration 5 Modeling Agent Behavior with Shipping Plans 5.1 Model 5.2 Methods of Shipping Plans 5.3 Implementing Task Acceptance Strategies with Shipping Plans 6 Simulation Experiments 6.1 BiciMAD GPS Dataset 6.2 Simulation Setup 6.3 Simulation Results 7 Conclusion References EUMAS 2020 Session 4: Agent-Based Models, Social Choice, Argumentation, Model-Checking Measuring the Strength of Rhetorical Arguments 1 Introduction 2 Knowledge Representation and Negotiating Agents 3 Threats, Rewards, and Appeals 4 Strength Measurement Model 4.1 Pre-conditions: Credibility and Preferability 4.2 Steps of the Model 5 Empirical Evaluation 6 Related Work 7 Conclusions and Future Work References Understanding the Role of Values and Norms in Practical Reasoning 1 Introduction 2 Background 2.1 Preliminaries 2.2 Values 3 Integrating Values in Normative BDI Agents 3.1 Value Language 3.2 Value-Based Reasoning 3.3 Value Properties 4 Discussion References Predicting the Winners of Borda, Kemeny and Dodgson Elections with Supervised Machine Learning 1 Introduction 2 Background 2.1 Voting Theory 2.2 Machine Learning 3 Datasets and Preprocessing 4 Factorization of Profiles 4.1 Labeling Profiles 4.2 Representation 1 4.3 Representation 2 4.4 Representation 3 5 Experiments and Testing 5.1 Borda Results 5.2 Kemeny Results 5.3 Dodgson Results 6 Discussion 7 Related Work 8 Conclusions References From Virtual Worlds to Mirror Worlds: A Model and Platform for Building Agent-Based eXtended Realities 1 Introduction 2 Background and Related 2.1 Mirror Worlds 2.2 Features and Challenges 2.3 Related Work 3 A Model for Mirror Worlds 3.1 Structure 3.2 Mirror World General Laws 4 A Platform for Agent-Based XR Based on Mirror Worlds 4.1 Platform Architecture 4.2 The Platform Model and APIs 4.3 Evaluation in a Real Context 5 Concluding Remarks References Model-Checking Information Diffusion in Social Networks with PRISM 1 Introduction 2 Information Diffusion in Social Networks 3 PRISM Background and Theory 4 Model-Checking Infection Models 4.1 Classic SIS Model 4.2 Taking the Agent View: Informational State and Opinion Broadcast 5 Model-Checking Threshold Influence Models 6 Analysing Larger Networks 7 Related Work 8 Discussion 9 Summary References “Roads? Where We’re Going We Don’t Need Roads.” Using Agent-Based Modeling to Analyze the Economic Impact of Hyperloop Introduction on a Supply Chain 1 Introduction 2 Literature Analysis 2.1 Hyperloop 2.2 Impact of High-Speed Trains 3 Model Presentation 4 Model Analysis 4.1 Methods 4.2 Results 4.3 Discussion 5 Conclusions and Future Research References Sensitivity to Initial Conditions in Agent-Based Models 1 Introduction and Motivations 2 Models and Methods 2.1 A Simple Market Model 2.2 The Reynold’s Flocking Model 3 Methods 4 Results and Discussion 4.1 Simple Market Model 4.2 Reynolds’ Flocking Model 5 Conclusions, Limitations and Future Works References EUMAS 2020 Session 5: Agent-Oriented Software Engineering, Game Theory, Task Allocation, Learning Statecharts and Agent Technology: The Past and Future 1 Introduction 2 Background on Statecharts 3 Statecharts and Agent Oriented Software Engineering 4 Discussion 5 Future Directions 6 Conclusion References A Game of Double Agents: Repeated Stackelberg Games with Role Switch 1 Introduction 2 Background 2.1 Bayesian Stackelberg Game 2.2 Monte Carlo Tree Search (MCTS) 3 Proposed Methodology 3.1 Assumptions and Complexity Results 3.2 Solution Method: SAFE 4 Empirical Evaluation 4.1 Synthetic Games 4.2 Malware Analysis 5 Conclusions References Learning Summarised Messaging Through Mediated Differentiable Inter-Agent Learning 1 Introduction 2 Related Work 3 Background: DQN 4 Proposed Approach 4.1 M-DIAL Architecture 4.2 Training 5 Experimental Results 5.1 Hidden Reward 5.2 Pong 6 Conclusion References Integrating Deep Learning and Non-monotonic Logical Reasoning for Explainable Visual Question Answering 1 Introduction 2 Related Work 3 Architecture 3.1 Feature Extraction Using CNNs 3.2 Classification Using Non-monotonic Logical Reasoning or Decision Trees 3.3 Answering Explanatory Questions 3.4 Learning State Constraints 3.5 Planning with Domain Knowledge 4 Experimental Setup and Results 4.1 Experimental Results: VQA + Learn Axiom 4.2 Experimental Results: Learn Axiom + Plan 5 Discussion and Conclusions References Multiagent Task Coordination as Task Allocation Plus Task Responsibility 1 Introduction 2 Conceptual Analysis and Formal Preliminaries 2.1 Conceptual Analysis 2.2 Concurrent Epistemic Game Structures 3 Specification 4 Allocating Tasks in TasCore 5 Ascribing Responsibility in TasCore 6 Discussion and Concluding Remarks References Anytime and Efficient Coalition Formation with Spatial and Temporal Constraints 1 Introduction 2 Problem Formulation 2.1 Basic Definitions 2.2 Coalition Allocations 2.3 Coalition Values 2.4 Constraints 2.5 Objective Function 3 Coalition Formation with Improved Look-Ahead 3.1 The Concept of CFLA2 3.2 Phase 1: Defining the Legal Agent Allocations 3.3 Phase 2: Selecting the Best Coalition for Each Task 3.4 Phase 3: Defining the Degree of Each Task 3.5 Phase 4: Overall Procedure of CFLA2 3.6 Analysis and Discussion 4 Cluster-Based Task Scheduling 4.1 Selecting the Best Task for Each Agent 4.2 Overall Procedure of CTS 4.3 Analysis and Discussion 5 Comparison Tests 5.1 Setup 5.2 Results 6 Conclusions References Author Index