A web document clustering algorithm based on concept of neighbor

As the WWW devloped rapidly, it becomes the most important resource gradually that transfers and shares the global information as well as being full of the latent capacity. Recent years, the researches of the Web mining have been concerned broadly and gotten a great deal of achievements simultaneously. The nearest neighbor technique, which is hierarchical clustering method based on distance, has been applied to many cases widely for the efficiency and validity. In this paper, based on the Vector Space Model (VSM) of the Web documents, We improved the nearest neighbor method, put forward a new Web document clustering algorithm, and researched the validity and scalability of the algorithm, the time and space complexity of the algorithm.