期刊名称：International Journal of Computer Technology and Applications
电子版ISSN：2229-6093
出版年度：2012
卷号：3
期号：2
页码：578-582
出版社：Technopark Publications
摘要：Web search engine are often forced to pass through long ordered list of documents called snippets. Snippets are web document attributes. These snippets are returned by search engines. The basis of document clustering is an alternative method of organizing retrieval results. Clustering yet needed to be deployed for the search engines. The approach adopted is formulation, simulation; formulation refers to the decomposition of different page rank values. Improved data clustering kmeans algorithm performs better results. Purpose of adopted web mining approach is to preserve web page conceptually similar, in page rank, link structure mining and probabilistic hybrid approach. Final goal is to eliminate the problem of increasing accuracy also with speed. As a result search engine gain popularity with incorporation of web mining. Proposed method ought to produce desired quality of clusters with probabilistic hybrid approach. Most users are unwilling to wait while accurate results are required at user’s end with probabilistic approach. Proposed model should be incorporated with the search engines to gain optimized results in terms of accuracy and speed
关键词：Fuzzy Clustering & Fuzzy Merging;Single Value Decomposition; Probablostic Distribution Function; Iris Dataset;Meta Information; Uniform Data Function; High Dimensional Data
