Improvement of Apriori Algorithm by Reducing Number of Database Scans and Generation of Candidate Keys

Anneshya Ghosh, Ambar Dutta

Abstract


Finding frequent itemsets is a key step in various data mining applications to find interesting patterns from databases. Association rule mining is an important technique in data mining. Apriori algorithm is most basic, simplest and classical algorithm of association rule mining. This algorithm is considered as an efficient algorithm, but still it has some drawbacks. In the literature, there exist a number of improvements for mining association rules based on Apriori algorithm. According to the problem that the traditional Apriori algorithm needs to scan database frequently, an improved strategy and corresponding algorithm is put forward in this paper. A comparative study of the traditional Apriori, existing improvements and proposed improved version of Apriori algorithm is presented in this paper with the help of different databases.

Keywords: Data Mining; Apriori algorithm; Support; Database; Frequent Itemsets.


Full Text:

PDF


DOI: https://doi.org/10.26483/ijarcs.v7i2.2637

Refbacks

  • There are currently no refbacks.




Copyright (c) 2016 International Journal of Advanced Research in Computer Science