Coop CBR

Cooperation Among Case-based Reasoning Agents

"Communication and Learning in a Wide World"

Here at the IIIA, we at the investigating several modes of cooperation among homogeneous agents with learning capabilities in a framework called federated learning.

In this page we focus on the cooperation among agents that learn and solve problems using Case-based Reasoning (CBR). Currently we are investigating several modes of cooperation:

• DistCBR - Distributed Case-based Reasoning Agents
• ColCBR - Collective Case-based Reasoning Agents
• Proactive Learning - Learning Agent's Competence Models to coordinate tasks in multiagent systems
• Auction-based Retrieval - Market-based mechanisms for coordinating retrieval of cases in multiagent systems

New article on competence models

Knowledge and Experience Reuse through Communication among Competent (Peer) Agents

(Available online [compressed PDF])

Francisco Martín, Enric Plaza, and Josep Lluís Arcos.

This article addresses an extension of the knowledge modelling ap- proaches, namely to multi-agent systems where communication and coor- dination are necessary. We propose the notion of competent agent and dene the basic capabilities of these agents for the extension to be ef- fective. An agent iscompetent when it is able to reason about its own competence and that of the other agents with which it cooperates in a given domain. In our framework, an agent has competence models of itself and of its acquaintances from which it can decide, for a specic problem to be solved, the type of cooperative activity it can request and from which agent. In this paper we focus on societies of peer agents, i. e. agents that are able to solve the same type of task but that may have dierent degrees of competence for specific problem ranges.

Published in International Journal of Software Engineering and Knowledge Engineering, Vol. 9, No. 3, 319-341

Article on Cooperation among Case-Based Reasoning Agents

Cooperative Case-Based Reasoning

(Available online [compressed PDF] | [compressed PS])

Enric Plaza, Josep Lluís Arcos, and Francisco Martín.

We are investigating possible modes of cooperation among homogeneous agents with learning capabilities. In this paper we will be focused on agents that learn and solve problems using Case-based Reasoning (CBR), and we will present two modes of cooperation among them: Distributed Case-based Reasoning (DistCBR) and Collective Case-based Reasoning (ColCBR). We illustrate these modes with an application where different CBR agents able to recommend chromatography techniques for protein purification cooperate. The approach taken is to extend Noos, the representation language being used by the CBR agents. Noos is knowledge modeling framework designed to integrate learning methods and based on the task/method decomposition principle. The extension we present, Plural Noos, allows communication and cooperation among agents implemented in Noos by means of three basic constructs: alien references, foreign method evaluation, and mobile methods.

Published in G. Weiss (Ed.) Distributed Artificial Intelligence meest Machine Learning, Lecture Notes in Artificial Intelligence, Springer Verlag, num. 1221, pp.180-201. (1997)

Article on market-based mechanisms for distributed case retrieval

Auction-based retrieval

(Available online [compressed PS])

Francisco Martín, and Enric Plaza.

We present Auction-based Retrieval ABR an approach for distributed case retrieval based on the economic metaphor of auction and on our current research on agent-based electronic trading. We focus on agent-mediated systems where each agent is able to reason from a (privately owned) case-base, has own interests, and nonetheless it is able to cooperate with the other to solve new problems. In this situation (called CoopCBR) case retrieval has an added difficulty, namely the coordination of case retrieval processes from multiple case-bases. Market institutions like auctions offer precisely well-known and well-founded coordination mechanisms for situations were the participants have limited information and individual objectives and interests.

Published in Proceedings of the "2n Congrés Català d'Intel.ligència Artificial", Girona, Spain, October 25-27, pp. 136-145.

Article on Distributed Case Libraries

Corporate Memories as Distributed Case Libraries

(Available online [Compressed PS] [Compressed PDF])

M.V. Nagendra Prasad and Enric Plaza.

In this paper, we discuss how, viewing corporate memories as distributed case libraries can benefit from existing techniques for distributed case-based reasoning for resource discovery and exploitation of previous expertise. We present two techniques developed in the context of multi-agent case-based reasoning for accessing and exploiting past experience from corporate memory resources. The first approach, called Negotiated Retrieval, deals with retrieving and assembling ``case pieces'' from different resources in a corporate memory to form a good overall case. The second approach, based on Federated Peer Learning, deals with two modes of cooperation called DistCBR and ColCBR that let an agent exploit the experience and expertise of peer agents to achieve a local task.

[Multiagent Learning] [Agents Page] [Features Terms] [Federated Learning] [NoosAgents] [Team Members]

noos@iiia.csic.es

http://www.iiia.csic.es/Projects/FedLearn/CoopCBR.html