A survey of query recommendation techniques for data warehouse exploration.

Abstract

Lots of data are gathered in datawarehouses that are navigated and explored for analytical purposes. Only recently has the problem of recommend- ing a datawarehouse query to a datawarehouse user attracted attention. In this paper, we propose a simple formal framework for expressing datawarehouse query recommendations. We propose to see the problem of recommending a datawarehouse query for exploration purposes as a function computing a set of queries and associated ratings given a query log, a session, a user profile, a datawarehouse instance, and an expectation function. The rating computed indi- cates the usefulness of each query for a session. With this viewpoint, we review and categorize the few techniques that, to the best of our knowledge, have been proposed and we illustrate them with a case study.