Querying Inductive Databases via Logic-Based User-Defined Aggregates

We show how a logic-based database language can support the various steps of the KDD process by providing: a high degree of expressiveness, the ability to formalize the overall KDD process and the capability of separating the concerns between the specification level and the mapping to the underlying databases and datamining tools. We generalize the notion of Inductive Data Bases proposed in [4,12] to the case of Deductive Databases. In our proposal, deductive databases resemble relational databases while user defined aggregates provided by the deductive database language resemble the mining function and results. In the paper we concentrate on association rules and show how the mechanism of user defined aggregates allows to specify the mining evaluation functions and the returned patterns.