Tim Baarslag

GENIUS: AN INTEGRATED ENVIRONMENT FOR SUPPORTING THE DESIGN OF GENERIC AUTOMATED NEGOTIATORS

The design of automated negotiators has been the focus of abundant research in recent years. However, due to difficulties involved in creating generalized agents that can negotiate in several domains and against human counterparts, many automated negotiators are domain specific and their behavior cannot be generalized for other domains. Some of these difficulties arise from the differences inherent within the domains, the need to understand and learn negotiators’ diverse preferences concerning issues of the domain, and the different strategies negotiators can undertake. In this paper we present a system that enables alleviation of the difficulties in the design process of general automated negotiators termed Genius, a General Environment for Negotiation with Intelligent multi‐purpose Usage Simulation. With the constant introduction of new domains, e‐commerce and other applications, which require automated negotiations, generic automated negotiators encompass many benefits and advantages over agents that are designed for a specific domain. Based on experiments conducted with automated agents designed by human subjects using Genius we provide both quantitative and qualitative results to illustrate its efficacy. Finally, we also analyze a recent automated bilateral negotiators competition that was based on Genius. Our results show the advantages and underlying benefits of using Genius and how it can facilitate the design of general automated negotiators.

Acceptance conditions in automated negotiation

In every negotiation with a deadline, one of the negotiating parties has to accept an offer to avoid a break off. A break off is usually an undesirable outcome for both parties, therefore it is important that a negotiator employs a proficient mechanism to decide under which conditions to accept. When designing such conditions one is faced with the acceptance dilemma: accepting the current offer may be suboptimal, as better offers may still be presented. On the other hand, accepting too late may prevent an agreement from being reached, resulting in a break off with no gain for either party. Motivated by the challenges of bilateral negotiations between automated agents and by the results and insights of the automated negotiating agents competition (ANAC), we classify and compare state-of-the-art generic acceptance conditions. We focus on decoupled acceptance conditions, i.e. conditions that do not depend on the bidding strategy that is used. We performed extensive experiments to compare the performance of acceptance conditions in combination with a broad range of bidding strategies and negotiation domains. Furthermore we propose new acceptance conditions and we demonstrate that they outperform the other conditions that we study. In particular, it is shown that they outperform the standard acceptance condition of comparing the current offer with the offer the agent is ready to send out. We also provide insight in to why some conditions work better than others and investigate correlations between the properties of the negotiation environment and the efficacy of acceptance conditions.

The Fifth Automated Negotiating Agents Competition (ANAC 2014)

In May 2015, we organized the Sixth International Automated Negotiating Agents Competition (ANAC 2015) in conjunction with AAMAS 2015. ANAC is an international competition that challenges researchers to develop a successful automated negotiator for scenarios where there is incomplete information about the opponent. One of the goals of this competition is to help steer the research in the area of multi-issue negotiations, and to encourage the design of generic negotiating agents that are able to operate in a variety of scenarios. 24 teams from 9 different institutes competed in ANAC 2015. This chapter describes the participating agents and the setup of the tournament, including the different negotiation scenarios that were used in the competition. We report on the results of the qualifying and final round of the tournament.

Decoupling Negotiating Agents to Explore the Space of Negotiation Strategies

Every year, automated negotiation agents are improving on various domains. However, given a set of negotiation agents, current methods allow to determine which strategy is best in terms of utility, but not so much the reason of success. In order to study the performance of the individual elements of a negotiation strategy, we introduce an architecture that distinguishes three components which together constitute a negotiation strategy: the bidding strategy, the opponent model, and the acceptance condition. Our contribution to the field of bilateral negotiation is threefold: first, we show that existing state of the art agents are compatible with this architecture; second, as an application of our architecture, we systematically explore the space of possible strategies by recombining different strategy components; finally, we briefly review how the BOA architecture has been recently applied to evaluate the performance of strategy components and create novel negotiation strategies that outperform the state of the art.

Learning about the opponent in automated bilateral negotiation: a comprehensive survey of opponent modeling techniques

A negotiation between agents is typically an incomplete information game, where the agents initially do not know their opponent’s preferences or strategy. This poses a challenge, as efficient and effective negotiation requires the bidding agent to take the other’s wishes and future behavior into account when deciding on a proposal. Therefore, in order to reach better and earlier agreements, an agent can apply learning techniques to construct a model of the opponent. There is a mature body of research in negotiation that focuses on modeling the opponent, but there exists no recent survey of commonly used opponent modeling techniques. This work aims to advance and integrate knowledge of the field by providing a comprehensive survey of currently existing opponent models in a bilateral negotiation setting. We discuss all possible ways opponent modeling has been used to benefit agents so far, and we introduce a taxonomy of currently existing opponent models based on their underlying learning techniques. We also present techniques to measure the success of opponent models and provide guidelines for deciding on the appropriate performance measures for every opponent model type in our taxonomy.

Heuristic-Based Approaches for CP-Nets in Negotiation

CP-Nets have proven to be an effective representation for capturing preferences. However, their use in multiagent negotiation is not straightforward. The main reason for this is that CP-Nets capture partial ordering of preferences, whereas negotiating agents are required to compare any two outcomes based on the request and offers. This makes it necessary for agents to generate total orders from their CP-Nets. We have previously proposed a heuristic to generate total orders from a given CP-Net. This paper proposes another heuristic based on Borda count, applies it in negotiation, and compares its performance with the previous heuristic.

Effective acceptance conditions in real-time automated negotiation

In every negotiation with a deadline, one of the negotiating parties must accept an offer to avoid a break off. As a break off is usually an undesirable outcome for both parties, it is important that a negotiator employs a proficient mechanism to decide under which conditions to accept. When designing such conditions, one is faced with the acceptance dilemma: accepting the current offer may be suboptimal, as better offers may still be presented before time runs out. On the other hand, accepting too late may prevent an agreement from being reached, resulting in a break off with no gain for either party. Motivated by the challenges of bilateral negotiations between automated agents and by the results and insights of the automated negotiating agents competition (ANAC), we classify and compare state-of-the-art generic acceptance conditions. We perform extensive experiments to compare the performance of various acceptance conditions in combination with a broad range of bidding strategies and negotiation scenarios. Furthermore we propose new acceptance conditions and we demonstrate that they outperform the other conditions. We also provide insight into why some conditions work better than others and investigate correlations between the properties of the negotiation scenario and the efficacy of acceptance conditions.

Measuring the performance of online opponent models in automated bilateral negotiation

An important aim in bilateral negotiations is to achieve a win-win solution for both parties; therefore, a critical aspect of a negotiating agents success is its ability to take the opponents preferences into account. Every year, new negotiation agents are introduced with better learning techniques to model the opponent. Our main goal in this work is to evaluate and compare the performance of a selection of state-of-the-art online opponent modeling techniques in negotiation, and to determine under which circumstances they are beneficial in a real-time, online negotiation setting. Towards this end, we provide an overview of the factors influencing the quality of a model and we analyze how the performance of opponent models depends on the negotiation setting. This results in better insight into the performance of opponent models, and allows us to pinpoint well-performing opponent modeling techniques that did not receive much previous attention in literature.

Predicting the Performance of Opponent Models in Automated Negotiation

When two agents settle a mutual concern by negotiating with each other, they usually do not share their preferences so as to avoid exploitation. In such a setting, the agents may need to analyze each others behavior to make an estimation of the opponents preferences. This process of opponent modeling makes it possible to find a satisfying negotiation outcome for both parties. A large number of such opponent modeling techniques have already been introduced, together with different measures to assess their quality. The quality of an opponent model can be measured in two different ways: one is to use the agents performance as a benchmark for the models quality, the other is to directly evaluate its accuracy by using similarity measures. Both methods have been used extensively, and both have their distinct advantages and drawbacks. In this work we investigate the exact relation between the two, and we pinpoint the measures for accuracy that best predict performance gain. This leads us to new insights in how to construct an opponent model, and what we need to measure when optimizing performance.

A Tit for Tat Negotiation Strategy for Real-Time Bilateral Negotiations

We describe the strategy of our negotiating agent, Nice Tit for Tat Agent, which reached the finals of the 2011 Automated Negotiating Agent Competition. It uses a Tit for Tat strategy to select its offers in a negotiation, i.e.: initially it cooperates with its opponent, and in the following rounds of negotiation, it responds in kind to the opponent’s actions.We give an overview of how to implement such a Tit for Tat strategy and discuss its merits in the setting of closed bilateral negotiation.