Authors: Djenni Rezoug Nachida; Nader Fahima; Boumahdi Fatima
Addresses: National Computer Science Engineering School, ESI, Oued Smar, Algiers, Algeria ' National Computer Science Engineering School, ESI, Oued Smar, Algiers, Algeria ' National Computer Science Engineering School, ESI, Oued Smar, Algiers, Algeria
Abstract: Mobile online analyses processing (OLAP) system offers to decision makers the real-time and relevant analyses anywhere and at anytime. In order, to generate them, a mobile OLAP should not only use user preferences, but also exploits information about contextual situation (meeting, business travel, office work, or home work) where analyses are done. For instance, when generating analyses, a mobile OLAP could take into account whether the decision maker's contextual situation is a business travel (uses a device with limited resources) or an office work (uses a device with high capacities). For this end, we investigate in this paper to propose a mobile context-aware recommender system (MCARS for short) based on both user preference and context. But, unfortunately, the limited resources in the MCARS make reducing a context acquisition a necessary need. To achieve this goal, our system proposes: 1) a learned approach which generates relevant contextual factors (contextual factors shown to be important); 2) deduces a relationship between a context and user's preferences (called contextual preferences); 3) and finally recommends a set of analysis based on user's contextual preferences.
Keywords: relevant context; context-aware recommender system; CARS; K2; knowledge-based recommender system.
International Journal of Information and Communication Technology, 2018 Vol.13 No.2, pp.149 - 175
Accepted: 10 May 2015
Published online: 14 Mar 2018 *