Abstract
Clustering has always been an important research topic that plays an important role in a broad range of applications, from information retrieval to CRM (Customer Relationship Management). It can be described as the process of dividing or grouping a set of data into classes of similar objects. The main goal of this paper is to develop an enhanced fuzzy clustering algorithm which determines the optimal number of clusters based on the detection of the two maximum values to split the worst clusters. For this, we use a new method incrementing the number of clusters in the main loop of an iterative clustering algorithm. Experimental results and comparisons are given to illustrate the performance of the new splitting method compared to the iterative one.