Coherent rule mining using AIR

Abstract

In the field of data mining, there are multiple data mining algorithms. This paper focuses on the proposed algorithm which works on logical coherence between data items and mines positive as well as negative rules. The proposed system overcomes limitations of a priori algorithm like reducing need of support and minimizing need of domain expert to decide that support and confidence. Indexing mechanism and set theory is used to enhance the performance. After generating coherent rules, recall method is applied for pruning. Comparative study of a priori algorithm and proposed algorithm is also discussed.