A Novel and Improved Apriori Algorithm

Abstract:

Apriori algorithm is the classical algorithm in data mining association rules. Because the Apriori algorithm needs scan database for many times, it runs too slowly. In order to improve the running efficiency, this paper improves the Apriori algorithm based on the Apriori analysis. The improved idea is that it transforms the transaction database into corresponding 0-1 matrix. Whose each vector and subsequent vector does inner product operation to receive support. And comparing with the given minsupport, the rows and columns will be deleted if vector are less than the minsupport, so as to reduce the size of the rating matrix, improve the running speeding. Because the improved algorithm only needs to scan the database once when running, therefore the running speeding is more quickly. The experiment also shows that this improved algorithm is efficient and feasible.

• Periodical: Applied Mechanics and Materials (Volume 721)
• Pages: 543-546
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.721.543
• Online since: December 2014
• Authors: Dong Juan Gu, Lei Xia
• Keywords: Apriori Association Rules, Data Mining, Improved Apriori Algorithm

References

[1] Han J. W, Kamber M. Data Mining Concept and Technology. Beijing: China Machine Press, 2001, pp.102-120.

[2] Liu Y.Y. Apply on College Score Analysis of Data Miming Technology, Technology & Economy in Areas of Communications, Vol. 4, 2008, 65-67.

[3] Wen R, Li R.F. A Optimal Apriori Algorithm, Computer System & Application, Vol. 12 (2008) No. 1, pp.94-120.

[4] Liu W.Y. Improvement and Study of Apriori Algorithm Based on linked list, Computer Application & Software, Vol. 29 (2012) No. 5, pp.267-269.

[5] Yuan W. L, Zhen Cheng, Zhuo M.Q. A Improved Apriori Algorithm, Computer Technology and Development, Vol. 18 (2008) No. 5, pp.50-54.