Incremental associative classification on distributed databases

Abstract

Distributed Data Mining (DDM) which is a process of extracting knowledge from distributed data without integrating them in a common database. Due to its vast application in real world application distributed data mining has been a most familiar research interest. As the associative classification technique proved to be most efficient classifier compare to other classifiers we can found certain proposals in literature which can perform associative classification over distributed databases. Even after incremental data mining proved to be most optimized way to upgrade mined rules when new set of transaction added to database, there are lack of proposals which can perform incremental mining over distributed databases. Considering these issues the article presents incremental associative classification model over horizontally distributed databases. Experimental conducted using synthesized datasets has shown encouraging results.