Mining Maximal Frequent Item Sets

Abstract

Data mining or knowledge discovery in databases (KDD) is a collection of exploration techniques based on advanced analytical methods and tools for handling a large amount of information. Mining association rule is a main content of data mining research at present, and emphasizes particularly is finding the relation of different items in the database. How to generate frequent item sets is the key and core. It is an important aspect in improving mining algorithm that how to decrease item set candidates in order to generate frequent item set effectively. Efficient algorithms for mining frequent items etc are crucial for mining association rules. Most existing work focuses on mining all frequent item sets (FI). However, since any subset of a frequent item set also is frequent, it is sufficient to mine only the set of maximal frequent item sets (MFI). In this paper we study the performance of existing approach, Max-Miner, for mining maximal frequent item sets. We have also developed an algorithm, called M-fp. We also present experimental results which shows that our method outperforms the existing method Max-Miner.