An Efficient Fuzzy Weighted Association Rule Mining with Enhanced Hits Algorithm

Association rule mainly focuses on large transactional databases. In association rule mining all items are considered with equal weightage. But it is not suitable for all datasets. The weight should be considered based on the importance of the item. In our previous work HITS algorithm (Hyperlink Induced Topic Search) is used to find the weight of an item w-support is calculated for generating frequent item sets. In this work enhanced HITS is used to calculate the weightage of the items. The enhanced HITS update the weight value in online manner. Fuzzy logic approach is applied to improve the association rule mining. So the proposed fuzzy weighted association rule mining with enhanced HITS satisfies downward closure property which decreases computation time; uninteresting rules can be pruned because of assigning weights to items, which also reduce the execution time. This paper introduces enhanced HITS algorithm and compute weights to describe the importance of attribute with respect to users intuition and integrate the options into mining weighted fuzzy association rule algorithm. Most weighted association rule mining eliminates extra steps during rules generation. Experiments show that the proposed algorithm is capable of discovering new rules in an effective manner by obtaining high confidence results. The comparison between fuzzy weighted association rule mining with enhanced HITS and weighted association rule mining with enhanced HITS is experimentally evaluated with food mart dataset as shown in enhanced version outperforms the weighted association rule mining with enhanced HITS.