Clustering Based OLAP Query Log Mining And Recommendation

Online Analytical Processing(OLAP) enable users to analyze multidimensional data interactively from multiple perspectives. Business data warehouse often contains multiple topics, involving multiple dimensions and organized into different levels, with the increasing complexity of data warehouse, the analysis process is often tedious since the user may have no idea of what the forthcoming query should be.In this paper, a clustering based OLAP log mining and recommendation method was proposed which includes two steps. In the offline mining step, to address data sparseness, queries are summarized into classes by clustering with query distance function, then, from session data a class sequence suffix tree is constructed as the query recommending model. In the online query recommending step, a user’s class sequence is mapped into a sequence of classes, by looking up the sequence of classes in the suffix tree, the query recommendations can be generated effectively. Our approach has been implemented with the open source Mondrian OLAP server to recommend MDX queries and we have carried out some preliminary experiments that show its efficiency for generating effective query recommendations.