Abstract: By sharing knowledge resources via information technologies, a virtual organization (VO) enables member enterprises to share skills, costs, access to one another’s knowledges. Multi-agent systems (MAS) are capable of adapting themselves to unforeseen changes of their environment in autonomous manner, and differ from more traditional software systems because agents are intended to be independent autonomous, reactive, pro-active and sociable software entities.In this paper we present a multi-agent approach for handling virtual organization (VO) knowledge management operations. This research proposes to develop a multi-agent system using JADE (Java Agent Development Environment) for knowledge management in a virtual organization, considering agent characteristics, functionalities, communication, cooperation and coordination. Agents have the ability to reorganize themselves in order to bring closer agents capable of cooperating in knowledge problem. We are interested, in this paper, to adaptation in virtual organizations where agents share a same goal.
Abstract: The general structure of the paper is the following: We use the formalism of the Recursive Modelling Method for the purpose of decision theoretic calculations. The advantage of Recursive Modelling Method, when used for expected utility calculation, is that it is able to succinctly represent the content of the agents, including its preferences, abilities, and beliefs about the physical world, as well as the agents beliefs about the other agents, their preferences and abilities, their beliefs about the world and about other agents, their beliefs about others beliefs, and so on. We consider examples of various types of communicative acts. The results address the communicative acts that agents can use to share information about their environment (I call these modelling agents), acts used to express the current intention of the user (intentional messages), and acknowledging messages. We then present results on the agreement between the method of message selection and messages that humans choose, and show an experimental validation of our framework in a simulated multi-agent environment. Open problems and future work.
Notes: ISI Document Delivery No.: BHN63 Times Cited: 0 Cited Reference Count: 8 Popirlan, Claudiu Popirlan, Cristina Proceedings Paper 11th World Multi-Conference on Systemics, Cybernetics and Informatics/13th International Conference on Information Systems Analysis and Synthesis JUL 08-11, 2007 Orlando, FL Int Inst Informat & System 14269 LORD BARCLAY DR, ORLANDO, FL 32837 USA
