Pellucid Agent Architecture for Knowledge Management in Public Organizations
Institute of Informatics
Slovak Academy of Sciences, Bratislava, SR
e-mail:{laclavik.ui, balogh.ui, hluchy.ui}@savba.sk
Abstract
In this paper we describe Pellucid Agent Architecture for experience knowledge management. This architecture is suitable for public organizations, which enact administration processes. Pellucid captures knowledge from experienced employees, explores new knowledge and assists unexperienced employee by providing relevant knowledge. Pellucid spontaneously gives advices (recomended contacts, useful documents, etc.) to employees based on the work currently performed in scope of the process. Pellucid architecture is also suitable for many other knowledge management applications. Design and implementation of the architecture is presented. Pellucid is designed using OOP, Agent Oriented Programming (AOP) and Knowledge Management Technologies and Methodologies (UML, AUML and Protege). Pellucid is implemented in Java and JADE. Ontology is used for knowledge representation and for inter-agent communication. Knowledge is stored in Organizational Memory which is stored in object-oriented database.
1. Introduction
Pellucid (Platform for Organizationally Mobile Public Employees) is an EC-founded project (5FP RTD IST-2001-34519). The overall objective of Pellucid is to develop an adaptable platform for assisting organizationally mobile employees, in effect to re-engineer their work in the organization. This will improve organization effectiveness and efficiency by formalization, recording, storage and preservation of experience and knowledge; and support workers during integration in a new department or role by giving access to specific knowledge and experience accumulated in the past. At the technical level, the objective is to develop and integrate several advanced technologies in a customizable agent-based architecture. These technologies include autonomous cooperating agents; responsive interaction with the end-users; organizational memory; workflow and process modeling; and metadata for accessing document repositories. The objective is also to obtain experience with customization and to formulate guidelines for the best practice in using the Pellucid platform in assistance of organizationally mobile workers [1].
The main aim of the Pellucid System is to capture, record and capitalize the knowledge of current employees about their work in the organization. The stored data are successively analyzed and searched for knowledge and correlations between the used data and activities in working processes, in which the employee is taking part. The discovery of knowledge is performed by various means of data mining techniques as well as agent reasoning methodologies. Recent software for...