Mining Association Rule with Reducing Candidates Generation

Abstract

Summarizing customer reviews is one of the most important and recent technique in web content mining. Mining association rules is usually used in the treatments of this technique. Mainly, Apriori algorithm is the first mining association rules algorithm that pioneered the use of support-based pruning to control the exponential growth of candidate itemsets. This paper presents a new algorithm that is a modification of the Apriori algorithm to find all frequent nitemsets without generating useless candidates to enhancement the performance of summarizing customer reviews. This algorithm is found to be faster than the previous ones. It passes only once over the transaction dataset to get the frequent 1-itemsets which contains all information about the transaction dataset. Repeatedly, we directly generate k-itemsets from the previous (k-1)-itemsets. By using this algorithm there is no need to generate the candidates while generating k-itemsets. Finally, an overall system is introduced to apply our algorithm for constructing faster summarizing customer reviews.