Web Page Clustering Using Cemetery Organization Behavior Of Ants

Abstract

Clustering is the unsupervised classification of patterns (data items, observations or feature vectors) into groups (clusters). Clustering problem has been addressed by the researchers of many disciplines in different contexts. Due to the escalating amount of data available online, the World Wide Web has become one of the most precious resource for information retrievals and knowledge discoveries. Web mining technologies are the right solutions for knowledge discovery on the Web. In this paper, we focus on web page clustering based on their content. A web page clustering system is valuable in web search for grouping search results into strongly related sets of documents. It can improve similarity search by focusing on sets of pertinent documents. At the same time, as the large variety of noisy information is embedded in web pages, web page clustering is much more intricate than pure-text clustering. This paper addresses web page clustering problem through the technique inspired by cemetery organization behavior of ants. Technique proposed by us begins by reducing the dimensionality of index of web pages with the application of Latent Semantic Indexing (LSI). Web pages are then transformed to two dimensional grid space using cemetery organization behavior of ants. Web pages represented in this two dimensional grid space are finally clustered using k-means algorithm. Paper also demonstrates impact of dimensionality reduction by means of LSI and distance measure on web page clustering results is also demonstrated.