Susan E. Lander

A GENERIC MODEL FOR INTELLIGENT NEGOTIATING AGENTS

Research in Cooperative Distributed Problem Solving (CDPS) considers how problem-solving tasks should be allocated among a group of agents and how the agents should coordinate their actions to achieve effective problem solving. For some CDPS systems, negotiation plays an important role in how agents cooperate. We define negotiation as the process of information exchange by which the agents act to resolve inconsistent views and to reach agreement on how they should work together in order to cooperate effectively. We describe a generic model, the Recursive Negotiation Model (RNM), that can serve as a basis for classifying and specifying where conflict resolution among multiple experts, viewpoints, or types of reasoning is needed in building a sophisticated CDPS system. This model defines where and how negotiation can be applied during problem solving based on structuring problem solving into four stages: problem formulation, focus-of-attention, allocation of goals or tasks to agents, and achievement of goals or tasks. We further discuss how the degree of agent participation in control decisions, including decisions about assigning responsibility to agents, influences the nature of negotiation within a particular system. Through this model, we emphasize that negotiation may be a recursive, complex, and pervasive process that is used to resolve conflicts in both domain-level and control-level problem solving. Finally, we survey existing negotiation frameworks and how they relate to our generic model.

Retrieval and Reasoning in Distributed Case Bases1

Abstract The proliferation of electronically available networked information has led researchers to examine the issues involved in developing automated methods for gathering information in response to a query from a user. However, most of this literature deals with locating, gathering, and selecting the best response to a query from among a multitude of responses from different repositories or digital libraries. This paper deals with a different model of response to a query, involving composition of mutually related partial responses spread across a network of information repositories. We present a system for cooperative retrieval and composition of a case in which subcases are distributed across different agents in a multiagent system. From a Gestalt perspective, a good overall case may not be the one derived from the summation of best subcases. Each agents local view may result in best local cases, which when assembled may not result in the best overall case in terms of global measures. We propose a negotiation-driven case retrieval algorithm as an approach to dynamically resolving inconsistencies between different case pieces during the retrieval process.

Sharing metainformation to guide cooperative search among heterogeneous reusable agents

A reusable agent is a self-contained computational system that implements some specific expertise and that can be embedded into diverse applications requiring that expertise. Systems composed of heterogeneous reusable agents are potentially highly adaptable, maintainable, and affordable, assuming that integration issues such as information sharing, coordination, and conflict management can be effectively addressed. The authors investigate sharing metalevel search information to improve system performance, specifically with respect to how sharing affects the quality of solutions and the runtime efficiency of a reusable-agent system. They first give a formal description of shareable metainformation in systems where agents have private knowledge and databases and where agents are specifically intended to be reusable. They then present and analyze experimental results from a mechanical design system for steam condensers that demonstrate performance improvements related to information sharing and assimilation. Finally, they discuss the practical benefits and limitations of information sharing in application systems comprising heterogeneous reusable agents. Issues of pragmatic interest include determining what types of information can realistically be shared and determining when the costs of sharing outweigh the benefits.

Designing Integrated Engineering Environments: Blackboard-Based Integration of Design and Analysis Tools

Blackboard-based integration of multiple agents is a natural and viable technology for the implementation of concurrent-engineering (CE) environments This article presents a blackboard architecture that has been extended to support the integration of heterogeneous collabora tive agents The suitability of the architecture as an integration framework for CE applications is discussed, followed by a description of a proto type CE application developed as part of Ford Motor Companys participation in the RRM (Rapid Response Manufacturing) Consortium 1 The prototype is a mixed-initiative, multiuser, multiprocessing environment that integrates heterogeneous agents working on crankshaft design, analy sis, and manufacturing-feasibility assessment The effectiveness of the RRM prototype and the ease and speed with which it was created demon strates that blackboard-based integration technology is appropriate for building multiagent concurrent-engineering applications

Conflict Resolution Strategies for Cooperating Expert Agents

Problem-solving approaches which incorporate specialized cooperating expert agents seem intuitively appropriate for many complex problems. However, integrating diverse expertise requires that the experts have some mechanism for dealing with conflicts that occur during problem-solving. We describe the Cooperating Experts Framework (CEF), a framework developed to support cooperative problem-solving among sets of knowledge-based systems with limited information about each other’s local states. The systems solve subproblems relevant to their specific expertise and integrate their efforts using conflict resolution strategies that are appropriate to the problem solving context. In choosing a strategy CEF makes tradeoffs between the potential quality of a solution, the amount of processing required to apply a strategy, and the effect of local changes on the global solution. We also describe TEAM, a system implemented in the CEF framework, that designs steam condensers.

Learning organizational roles for negotiated search in a multiagent system

This paper presents studies in learning a form of organizational knowledge called organizational roles in a multi-agent agent system. It attempts to demonstrate the viability and utility of self-organization in an agent-based system involving complex interactions within the agent set. We present a multi-agent parametric design system called L-TEAM where a set of heterogeneous agents learn their organizational roles in negotiated search for mutually acceptable designs. We tested the system on a steam condenser design domain and empirically demonstrated its usefulness. L-TEAM produced better results than its non-learning predecessor, TEAM, which required elaborate knowledge engineering to hand-code organizational roles for its agent set. In addition, we discuss experiments with L-TEAM that highlight the importance of certain learning issues in multi-agent systems.

Knowledge-based conflict resolution for cooperation among expert agents

Cooperating human experts are able to integrate their skills and knowledge productively to achieve goals beyond their individual capabilities. Machine agents may someday increase their power similarly by working in teams of specialized experts. To do this, the systems must be able to communicate knowledge, propose solutions, resolve conflicts that occur during problem-solving, and agree on results. We describe the Cooperating Experts Framework (CEF), a generic framework that supports cooperative problem-solving among sets of knowledge-based systems. The systems solve subproblems relevant to their specific expertise and integrate their efforts using conflict resolution strategies. CEF provides scheduling and communication support for the agents, a set of conflict resolution strategies, and a set of heuristics for choosing the most effective strategy for the situation. We also describe STEAMER, a system implemented in the CEF framework, that designs steam condensers.

A framework for the integration of cooperative knowledge-based systems

Solving a complex problem often requires the application of knowledge from multiple experts; however, integrating diverse expertise requires that the experts be able to resolve their conflicting goals and beliefs with respect to the salient issues. The authors describe the cooperating experts framework (CEF), a generic framework developed to support cooperative problem solving among sets of knowledge-based systems. The systems solve subproblems relevant to their specific expertise and integrate their efforts through negotiation. CEF provides two complementary styles of negotiation and a set of heuristics based on problem characteristics which are used to choose between them.<<ETX>>

The role of Learning in systems of reusable heterogeneous design agents

In this abstract, we discuss the use of learning techniques to improve performance and solution quality in multiagent parametric design. We have implemented the L-TEAM testbed for empirical evaluation of two forms of learning (described in detail below):