Affiliations: [a] Normandie Université, INSA Rouen, LITIS, Avenue de l’Université, Saint-Étienne-du-Rouvray cedex, France
| [b] Faculty of Management and Economics, The Technical University of Gdansk, Narutowicza 11/12, 80-233 Gdansk, Poland
Correspondence:
[*]
Corresponding author. Cecilia Zanni-Merk, Normandie Université, INSA Rouen, LITIS, Avenue de l’Université, BP 8, 76801 Saint-Étienne-du-Rouvray cedex, France. E-mail: [email protected].
Abstract: This article presents a survey on the use of KREM, a generic knowledge-based framework for building collective intelligence through experience. After a discussion on the disadvantages of the traditional architecture used to deploy intelligent systems, the KREM architecture (Knowledge, Rules, Experience, Meta-Knowledge) is presented. The novelty of the proposal comes from the inclusion of the capitalisation of experience and the use of meta-knowledge in the traditional architecture previously discussed. KREM improves the efficiency of traditional intelligent systems by allowing incomplete expert knowledge models to be used, gradually completing them, learning with experience. In addition, the use of meta-knowledge can guide their execution more effectively. This framework has been successfully used in various projects in different application areas, which are presented and discussed.
