Disjoint Web Document Clustering and Management in Electronic Commerce
Mei‐Ling Shyu, Shu‐Ching Chen, Choochart Haruechaiyasak, Chi‐Min Shu, Sheng-Tun Li · 2001
Due to the recent trend in electronic commerce, many companies provide their product related or any usable information on their Web sites for customer convenience. This company information is organized as separate Uniform Resource Locator (URL) pages. Each URL represents a Web document that can be linked to or from other documents via the hyperlinks. How to customize a company's Web site for its Web site layout so that it can target its potential customers to improve profits is important in electronic commerce. In this paper, we propose the Markov Model Mediator (MMM) mechanism to organize and manage the groups of related URLs into disjoint clusters for document management in electronic commerce. An experiment is conducted using a real data set and the experimental result shows that our proposed approach yields a better performance over all different tested cluster sizes in comparison with depth-first search (DFS), breadth-first search (BFS), and the random clustering strategies.