An improved incremental algorithm for mining weighted class-association rules

Abstract

Constructing fast and accurate classifiers for large data sets is an important task in data mining. Associative classification can produce more efficient and accurate classifiers than traditional classification techniques. Weighted class association rule (WCAR) mining reflects significance of items by considering their weight. Moreover, real time databases are dynamic. This influences the need for incremental approach for classification. Existing incremental classification algorithms suffer from issues like longer execution time and higher memory usage. This paper proposes an algorithm which uses hash structure to store weighted frequent items and the concept of difference of object identifiers to compute the support faster. For mining incremental databases, pre-large concept is used to reduce the number of re-scans over the original database. The proposed algorithm was implemented and tested on experimental data sets taken from UCI repository. The results show that proposed algorithm for mining WCARs gives better results compared to existing algorithm.