Agent-Mediated Electronic Commerce and Trading Agent Design and Analysis - AAMAS Workshop, AMEC 2008, Estoril, Portugal, May 12-16, 2008, and AAAI Workshop, TADA 2008, Chicago, IL, USA, July 14, 208, Revised, Selected Papers

Preface

Organization

Table of Contents

Preventing Under-Reporting in Social Task Allocation

Introduction

Preliminaries

An Exact VCG Mechanism for STAP

A Greedy Mechanism for STAP

A Greedy Allocation Algorithm

A Mechanism That Is Truthful with Respect to Under-Reporting

Experiments

Discussion and Conclusions

References

Reasoning and Negotiating with Complex Preferences Using CP-Nets

Introduction

Representing and Ordering Preferences

Ceteris Paribus Nets (CP-Nets)

Obtaining Preferences

Strategies for Generating Requests

Sequential Search Strategy

Depth Limited Search Strategy

Binary Search Strategy

Upper Random Strategy

Producer’s Strategy for Counter Offer

Experiments

Discussion

References

Using Priced Options to Solve the Exposure Problem in Sequential Auctions

Introduction

Options: Basic Definition

RelatedWork

Outline and Contribution of Our Approach

Expected Profit for a Sequence of n Auctions and 1 Synergy Bidder

Profit with n Unique Goods without Options

Profit with n Unique Goods with Options

When Options Can Benefit Both Synergy Bidder and Seller

The Case When Agents Are Better Off with Options

Synergy Bidder’s Profit-Maximizing Bid

Simulation of a Market with a Single Synergy Bidder

Synergy Bidder’s Strategy

Experimental Results: Market Entry Effect for One Synergy Bidder

Multiple Synergy Bidders

Two Synergy Bidders Interacting Indirectly through the Exercise Price Level

Direct Synergy Bidder Competition in the Same Auctions

Larger Simulation with Random Synergy Bidders’ Market Entry

Discussion and Further Work

References

Towards a Quality Assessment Method for Learning Preference Profiles in Negotiation

Introduction

Related Work and Problem Description

Quality Assessment Method

Quality Measures

Negotiation Domains and Profiles

Negotiation Strategies of the Opponent

Application and Experimental Results

Experimental Setup

Evaluation

Conclusion and Discussion

References

Using a Memory Test to Limit a User to One Account

Introduction

TestSpecifics

A Small Study with Human Subjects

A Game-Theoretic Analysis

Conclusions and Future Research

References

Appendix: Another Test Based on Recognizing Faces

Multi-attribute Regret-Based Dynamic Pricing

Introduction

Dynamic Pricing over Multiple Product Attributes

Minimax Regret-Based Algorithms

Minimax Regret-Based Attribute Prediction

Regret-Based Dynamic Pricing

Simulation Results

Comparison Strategies

Experimental Results

Related Work

Conclusion

References

On the Economic Effects of Competition between Double Auction Markets

Introduction

Background

Experimental Setup

Software

Traders

Markets

Experiments

Measurements

Results

Discussion

Conclusions

References

A Multiagent Recommender System withTask-Based Agent Specialization

Introduction

Related Work

Recommender Systems

Multiagent Recommender Systems

MAS Recommendation Model

The Agents

The Recommendation Algorithm

Experimental Results

Conclusions

References

Towards Automated Bargaining in Electronic Markets: A Partially Two-Sided Competition Model

Introduction

Alternating-Offers Bargaining with Agents’ Deadlines

The Proposed Model

Equilibrium Strategies

Base Case: One Buyer and One Seller

One-Sided Competition I: One Buyer and More Sellers

One-Sided Competition II: More Buyers and One Seller

Two-Sided Competition: More Buyers and More Sellers

Conclusions and Future Works

References

Bidding Heuristics for Simultaneous Auctions:Lessons from TAC Travel

Introduction

TACTravelGame

Bidding Heuristics

Marginal-Utility-Based Heuristics

Sample Average Approximation

Experiments in TAC Travel-Like Auctions

Decision-Theoretic Experiments with Perfect Distributional Prediction

Setup

Results

Decision-Theoretic Experiments with Imperfect Distributional Prediction

Setup

Results

Experiments with Competitive Equilibrium Prices

Setup

Decision-Theoretic Experiments

Game-Theoretic Experiments

Summary and Discussion of Experimental Results

Related Work

Conclusion

References

Applications of Classifying Bidding Strategies for the CAT Tournament

Introduction

Bidding Strategies

Classifying Traders by Bidding Strategies

Data Collection

Classification Using a Support Vector Machine

Classification Using a Hidden Markov Model

Alternative Classification Techniques

Utilizing Classification to Determine Optimal Action Policies

Experimental Results

Testing Environment

Fee Adjustments

Experiments

Determining Optimal Actions Using an MDP

Conclusion and Future Work

References

Coordinating Decisions in a Supply-Chain Trading Agent

Introduction

Overview of the TAC SCMGame

Agent Decision Processes

Game Balance

Agent Design and the Decision Coordination Problem

Predicted Sales Volume

Future Production Schedule

InventoryManagement

Central StrategyModule

Separate Supply and Demand Models

Internal Markets

Conclusions and Future Work

References

The 2007 TAC SCM Prediction Challenge

Introduction

The Prediction Challenge

Prediction Methods

Results and Analysis

Results

Average Daily Errors

Differences between Participants across Games

Differences between Participants across Days

Conclusion

References

Author Index