Knowledge datawarehouse: Web usage OLAP application

Abstract

Generally, OLAP analysis are based on both the observed data and a set of OLAP operators for restructuration and granularity modification. The goal is to discover patterns hidden into data. Unfortunately, this approach is also based on the analyst background. This latter assumes hypothesis according to his background and analyses data consequently: "hypothesis driven analysis". The integration of knowledge into data warehouse conduce to enriched analysis context where objects and their relations are explicitly represented, handled and visualized. We investigate a deep integration where the basic datawarehouse's operators consider both data and knowledge. This paper applies knowledge datawarehouse concept to Web usage analysis.