Université de Toulouse
It has recently been pointed out the necessity to come up with flexible, powerful means for analyzing past issued queries, usually stored in the DB query log, and decompose, store and handle them in a dedicated subsystem in order to better support any decisional task with the knowledge captured in these queries. This paper proposes a manipulation language tailored for OLAP query logs, stemming from the relational algebra. This language is based on binary relations over sequences of queries (called sessions). The authors propose two such relations allowing to group and order sessions. Examples of expressions in this language illustrate its interest for various user-centric approaches, like query recommendation or log summarization, particularly relevant in the context of data warehouse exploration.