Home > Published Issues > 2012 > Volume 3, No. 1, February 2012 >

Discovery of Scalable Association Rules from Large Set of Multidimensional Quantitative Datasets

Tamanna Siddiqui, M Afshar Aalam, and Sapna Jain
Jamia Hamdard/Department of Computer Science, Haryana, India

Abstract— In proposed approach, we introduce the problem of mining association rules in large relational tables containing both quantitative and categorical attributes. We have proposed an algorithm for Discovery of Scalable Association Rules from large set of multidimensional quantitative datasets using k-means clustering method based on the range of the attributes in the rules and Equidepth partitioning using scale k-means for obtaining better association rules with high support and confidence. The discretization process is used to create intervals of values for every one of the attributes in order to generate the association rules. The result of the proposed algorithm discover association rules with high confidence and support in representing relevant patterns between project attributes using the scalable k-means .The experimental studies of proposed algorithm have been done and obtain results are quite encouraging.

Index Terms— Data Mining, Association rules, k-means clustering, CBA tool, Discretization, Partitioning.

Cite: Tamanna Siddiqui, M Afshar Aalam, and Sapna Jain, "Discovery of Scalable Association Rules from Large Set of Multidimensional Quantitative Datasets," Journal of Advances in Information Technology, Vol. 3, No. 1, pp. 69-76, February, 2012.doi:10.4304/jait.3.1.69-76