TLNL: A novel two level node labeling algorithm for frequent pattern trees

Abstract

In the area of Data Mining, We generally use many techniques for data analysis, among them, association rule learning is a well-liked and well researched technique for discover the interesting relations among the variables in large databases. Association rules are a part of intelligent systems because all the intelligent systems are using the associations. Association rules are usually needed to satisfy a user-individual minimum support and minimum confidence at the same time. Apriori algorithm (Static) and FP_Growth(Dynamic) algorithms are the traditional algorithms used to extract the frequent itemsl. The Frequent Pattern-Growth algorithm is completely depends on fp-tree. In previous, the fp-tree node is labeled only with its support count, due to this, more time takes while traversing to extract the associated items with that particular item. In this paper we are more concentrated on the node labeling scheme of fp-tree in FP-Growth algorithm. Here we propose a new two level node labeling (TLNL) approach for frequent pattern growth tree. The proposed algorithms are fast and efficient algorithms. This paper overcomes the major inconveniences of FP-Growth algorithm for association rule mining with using the newly proposed approach.