OLAP queries context-aware recommender system

Abstract

It becomes hard and tedious to easily obtain relevant decisional data in large data warehouses. In order to ease user exploration during on-line analytical processing analysis, recommender systems are developed. However some recommendations can be inappropriate (irrelevant queries or non-computable queries). To overcome these mismatches, we propose to integrate contextual data into the recommender system. In this paper, we provide (i) an indicator of obsolescence for OLAP queries and (ii) a context-aware recommender system based on a contextual post-filtering for OLAP queries.