ICE - Intelligent Clustering Engine: A clustering gadget for Google Desktop
Expert Systems with Applications: An International Journal
Hi-index | 0.00 |
Recently various clustering approaches have been developed for web pages clustering optimization. Most of them take the vector model as their free-text analytical foundation. However these algorithms cannot perform well on problems involving many Ecommerce information-clustering objectives. A novel approach based on the EHM vector space model FCM clustering algorithm is proposed to deal with the problems in this paper. By introducing Ecommerce Hierachical Model(EHM), the Automatic Constructing Concept (ACC) algorithm is proposed at first. Through the ACC algorithm and fields keywords table, the Ecommerce concept objects are established automatically. The EHM-Based Fuzzy (EFCM) clustering is used to divide web pages into the different concept subsets. The experiment has compared it with such methods as Kmeans, Kmedoid, Gath-Geva clustering algorithm, and results demonstrate the validity of the new algorithm. According to classification performance, the EFCM algorithm shows that it can be clustering method for the semantic information searching in Internet.