Improving Skills Management using Objectives within a Multi-Agent System

Whatever the professional sector of activity, achievements are the results of dedicated processes and skill usage. If a project is successful, the knowledge an d know-how that have been used have to be capitalized for a po ssible use in other projects. However, this is a dynamic process. Skills are always evolving. The issue is how to improve the ma nagement of skills to make them more efficient over time. A new approach is proposed to address this problem. It is based on the management of objectives through a multi agent learning system where skills are represented as agents. Bein g autonomous entities, and having their own learning abilities, skills will enforce project efficiency, capitalizin g on past iterations. The key point of our approach is the ap propriate exploitation of past objectives to help the user id entifying the required skills for the new project.