Abstract—With the rapid development of clustering analysis
technology, there have been many application-specific
clustering algorithms, such as text clustering. K-Means
algorithm, as one of the classic algorithms of clustering
algorithms, and a textual document clustering algorithms
commonly used in the analysis process, is widely used because
of its simple and low complexity. This article in view of two big
limitations that the K-Means algorithm has, namely
requirements that users give the anticipated variety beforehand
integer K and random selection of initial variety center,
proposed K-Means improved algorithm based on the
association rules technology. This method proposed the concept
of the smallest rule covering set .It has relieved these two big
limitations of K-Means algorithm effectively. It is used for the
audit monitor target discovery and extraction process in social
security domain basic old-age insurance audit methods. Thus it
can provide better reference value and guiding sense for
auditors.
Index Terms—Clustering, text clustering, association rules,
K-means algorithm, monitoring indicators.
The authors are with the College of Computer Science and Technology at
Harbin Engineering University, China. (e-mail: liugang@hrbeu.edu.cn,
huangshaobin@hrbeu.edu.cn, s311060106@hrbeu.edu.cn,
duyudan@hrbeu.edu.cn).
[PDF]
Cite:Gang Liu, Shaobin Huang, Caixia Lu, and Yudan Du, "An improved K-Means Algorithm Based on Association Rules," International Journal of Computer Theory and Engineering vol. 6, no. 2, pp. 146-149, 2014.
