Set Representation of Itemset for Candidate Generation with Binary Search Technique

摘要

Till date, there are many association mining algorithms which have been used in varied applications. The performance of these mining algorithms are greatly affected by the execution time and the way the itemsets are stored in the memory. Most of the time is wasted while scanning the database and searching for the frequent itemsets. The search space keeps on increasing when the number of attributes in the database is large, especially when dealing with millions of transactions. Furthermore, the representation of itemsets in the memory plays a vital role when large databases are handled. To solve this problem, this paper shows an improvement of the algorithms by representing a concise representation and reducing the searching time for the frequent itemsets.