Onn Shehory

Coalition structure generation with worst case guarantees

Coalition formation is a key topic in multiagent systems. One may prefer a coalition structure that maximizes the sum of the values of the coalitions, but often the number of coalition structures is too large to allow exhaustive search for the optimal one. Furthermore, finding the optimal coalition structure is NP-complete. But then, can the coalition structure found via a partial search be guaranteed to be within a bound from optimum? We show that none of the previous coalition structure generation algorithms can establish any bound because they search fewer nodes than a threshold that we show necessary for establishing a bound. We present an algorithm that establishes a tight bound within this minimal amount of search, and show that any other algorithm would have to search strictly more. The fraction of nodes needed to be searched approaches zero as the number of agents grows. If additional time remains, our anytime algorithm searches further, and establishes a progressively lower tight bound. Surprisingly, just searching one more node drops the bound in half. As desired, our algorithm lowers the bound rapidly early on, and exhibits diminishing returns to computation. It also significantly outperforms its obvious contenders. Finally, we show how to distribute the desired search across self-interested manipulative agents.

Coalition formation with uncertain heterogeneous information

Coalition formation methods allow agents to join together and are thus necessary in cases where tasks can only be performed cooperatively by groups. This is the case in the Request For Proposal (RFP) domain, where some requester business agent issues an RFP - a complex task comprised of sub-tasks - and several service provider agents need to join together to address this RFP. In such environments the value of the RFP may be common knowledge, however the costs that an agent incurs for performing a specific sub-task are unknown to other agents. Additionally, time for addressing RFPs is limited. These constraints make it hard to apply traditional coalition formation mechanisms, since those assume complete information, and time constraints are of lesser significance there.To address this problem, we have developed a protocol that enables agents to negotiate and form coalitions, and provide them with simple heuristics for choosing coalition partners. The protocol and the heuristics allow the agents to form coalitions in the face of time constraints and incomplete information. The overall payoff of agents using our heuristics is very close to an experimentally measured optimal value, as our extensive experimental evaluation shows.

Coalition formation for large-scale electronic markets

Coalition formation is a desirable behavior in a multiagent system, when a group of agents can perform a task more efficiently than any single agent can. Computational and communications complexity of traditional approaches to coalition formation, e.g., through negotiation, make them impractical for large systems. We propose an alternative, physics-motivated mechanism for coalition formation that treats agents as randomly moving, locally interacting entities. A new coalition may form when two agents encounter one another and it may grow when a single agent encounters it. Such agent-level behavior leads to a macroscopic model that describes how the number and distribution of coalitions change with time. We increase the generality and complexity of the model by letting the agents leave coalitions with some probability. The model is expressed mathematically as a series of differential equations. These equations have steady state solutions that describe the equilibrium distribution of coalitions. Within a context of a specific multi-agent application, we analyze and discuss the connection between the global system utility the parameters of the model.

A Framework for Evaluating Agent-Oriented Methodologies

Multiple agent-oriented methodologies has been introduced in recent years, although only partial evaluations of these have been offered. As a result, it is difficult to select a methodology for a specific project. Additionally, there are no means for determining what the advantages and drawbacks of each methodology are. To resolve these problems, we devise a framework for evalu- ating and comparing agent-oriented methodologies. This framework focuses on four major aspects of a methodology: concepts and properties, notations and modelling techniques, process and pragmatics. We demonstrate the usage of the suggested framework by evaluating the GAIA methodology. This evaluation identifies the strengths and the weaknesses of GAIA, thus exemplifying the ca- pabilities of our framework.

Multi-Agent Coordination through Coalition Formation

Incorporating coalition formation algorithms into agent systems shall be advantageous due to the consequent increase in the overall quality of task performance. Coalition formation was addressed in game theory, however the game theoretic approach is centralized and computationally intractable. Recent work in DAI has resulted in distributed algorithms with computational tractability. This paper addresses the implementation of distributed coalition formation algorithms within a real-world multi-agent system. We present the problems that arise when attempting to utilize the theoretical coalition formation algorithms for a real-world system, demonstrate how some of their restrictive assumptions can be relaxed, and discuss the resulting benefits. In addition, we analyze the modifications, the complexity and the quality of the cooperation mechanisms. The task domain of our multi-agent system is information gathering, filtering and decision support within the WWW.

Agent cloning: an approach to agent mobility and resource allocation

Multi-agent systems are subject to performance bottlenecks in cases where agents cannot perform tasks by themselves due to insufficient resources. Solutions to such problems include passing tasks to others or agent migration to remote hosts. We propose agent cloning as a more comprehensive approach to the problem of local agent overloads. Agent cloning subsumes task transfer and agent mobility. According to our paradigm, agents may clone, pass tasks to others, die, or merge. We discuss the requirements of implementing a cloning mechanism and its benefits in a multi-agent system, and support our claims with simulation results.

Feasible Formation of Coalitions Among Autonomous Agents in Nonsuperadditive Environments

Cooperating and sharing resources by creating coalitions of agents are important ways for autonomous agents to execute tasks and to maximize payoff. Such coalitions will form only if each member of a coalition gains more by joining the coalition than it could gain otherwise. There are several ways of creating such coalitions and dividing the joint payoff among the members. In this paper we present algorithms for coalition formation and payoff distribution in nonsuperadditive environments. We focus on a low‐complexity kernel‐oriented coalition formation algorithm. The properties of this algorithm were examined via simulations. These have shown that the model increases the benefits of the agents within a reasonable time period, and more coalition formations provide more benefits to the agents.

A Planning Component for RETSINA Agents

In the RETSINA multi-agent system, each agent is provided with an internal planning component—HITaP. Each agent, using its internal planner, formulates detailed plans and executes them to achieve local and global goals. Knowledge of the domain is distributed among the agents, therefore each agent has only partial knowledge of the state of the world. Furthermore, the domain changes dynamically, therefore the knowledge available might become obsolete.

The Advantages of Compromising in Coalition Formation with Incomplete Information

This paper presents protocols and strategies for coalition formation with incomplete information under time constraints. It focuses on strategies for coalition members to distribute revenues amongst themselves. Such strategies should preferably be stable, lead to a fair distribution, and maximize the social welfare of the agents. These properties are only partially supported by existing coalition formation mechanisms. In particular, stability and the maximization of social welfare are supported only in the case of complete information, and only at a high computational complexity. Recent studies on coalition formation with incomplete and uncertain information address revenue distribution in a naïve manner. In this study we specifically refer to environments with limited computational resources and incomplete information. We propose a variety of strategies for revenue distribution, including the strategy in which the agents attempt to distribute the estimated net value of a coalition equally. A variation of the equal distribution strategy in which agents compromise and agree to a payoff lower than their estimated equal share, was specifically examined. Our experimental results show that, under time constraints, the compromise strategy is stable and increases the social welfare compared to non-compromise strategies.

A Scalable Agent Location Mechanism

Large scale open multi-agent systems where agents need services of other agents but may not know their contact information require agent location mechanisms. Solutions to this problem are usually based on middle-ware such as matchmakers, brokers, yellow-pages agents and other middle agents. The disadvantage of these is that they impose infrastructure, protocol and communication overheads, and they do not easily scale up. We suggest a new approach to agent location, which does not require middle agents and protocols for using them. Our approach is simple and scales up with no infrastructure or protocol overheads, thus may be very useful for large scale MAS. In this paper, we analytically study the properties of our approach and discuss its advantages.