Author : Sonali Bodekar 1
Date of Publication :7th November 2016
Abstract: The Web is an abundant source of data mining which is fast growing and dynamic that provides ample opportunities which are often not used. Due to its huge amount and the unstructured nature web data represent a real challenge to traditional data mining techniques. The constantly increasing demand of finding pattern from large data enhances the association rule mining. The traditional algorithm for association rule discovery is Apriori. Scanning the database many times is the drawback of Apriori algorithm, so that it doesn’t work well with the large database. Researchers developed a plenty of algorithms and techniques for finding association rules. The generation of candidate set is the main problem. Among the existing techniques, the most efficient and scalable approach is frequent pattern growth (FP-growth) method. Generation of a massive number of conditional FP tree is the main obstacle of FP growth. In this research paper, we proposed an algorithm improved FP tree with a table for mining association rules. This algorithm mines all possible frequent item set without conditional FP tree generation. Our proposed method implemented the improved FP-Tree based on MapReduce framework which has high achieving performance compared with the basic FP-Growth. It also gives the frequency of frequent items to evaluate the desired association rule and enhance the time efficiency of mining association rule.
Reference :