Logic-based association rule mining in XML documents

Abstract

In this paper, we propose a new framework, called XLogic-Miner, to mine association rules from XML data. We consider the generate-and-test and the frequent-pattern growth approaches. In XLogic-Miner, we propose an novel method to represent a frequent-pattern tree in an objectrelational table and exploit a new join operator developed in the paper. The principal focus of this research is to demonstrate that association rule mining can be expressed in an extended datalog program and be able to mine XML data in a declarative way. We also consider some optimization and performance issues. © Springer-Verlag Berlin Heidelberg 2006.