Nicolas Spyratos

Towards mining frequent queries in star schemes

The problem of mining all frequent queries in a database is intractable, even if we consider conjunctive queries only. In this paper, we study this problem under reasonable restrictions on the database, namely: (i) the database scheme is a star scheme; (ii) the data in the database satisfies a set of functional dependencies and a set of referential constraints. We note that star schemes are considered to be the most appropriate schemes for data warehouses, while functional dependencies and referential constraints are the most common constraints that one encounters in real databases. Our approach is based on the weak instance semantics of databases and considers the class of selection-projection queries over weak instances. In such a context, we show that frequent queries can be mined using level-wise algorithms such as Apriori.

Rewriting aggregate queries using functional dependencies

In this paper we consider the problem of answering OLAP queries in the context of a data warehouse using materialized views. Materialized views are stored in a cache and our approach uses partition semantics to rewrite incoming queries in terms of the stored views. We take advantage of functional dependencies between data to derive efficient rewriting algorithms and to define a cache maintenance method that optimizes the storage required for the materialized views.

TIGER: Querying large tables through criteria extension

Sales on the Internet have increased significantly during the last decade, and so, it is crucial for companies to retain customers on their web site. Among all strategies towards this goal, providing customers with a flexible search tool is a crucial issue. In this paper, we propose an approach, called TIGER, for handling such flexibility automatically. More precisely, if the search criteria of a given query to a relational table or a Web catalog are too restrictive, our approach computes a new query combining extensions of the criteria. This new query maximizes the quality of the answer, while being as close as possible to the original query. Experiments show that our approach improves the quality of queries, in the sense explained just above.