Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
552823 | Decision Support Systems | 2009 | 9 Pages |
Abstract
World Wide Web is a huge information space, making it a valuable resource for decision making. However, it should be effectively managed for such a purpose. One important management technique is clustering the web data. In this paper, we propose some developments in clustering methods to achieve higher qualities. At first we study a new density based method adapted for hierarchical clustering of web documents. Then utilizing the hyperlink structure of web, we propose a new method that incorporates density concepts with web graph. These algorithms have the preference of low complexity and as experimental results reveal, the resultant clusters have high quality.
Related Topics
Physical Sciences and Engineering
Computer Science
Information Systems
Authors
Morteza Haghir Chehreghani, Hassan Abolhassani, Mostafa Haghir Chehreghani,