An Efficient Web Content Mining Using Divide and Conquer Approach

In Context of web mining, large collection of web documents are used in the process of mining to extract more useful information. Most of the web information is irrelevant. Web document presents 10-15% of data using 85-90% of tags. The previous researchers on web mining proposed many methods, for mining web documents, but all these mining methods process documents without consider size of the document. Let N Documents are used in the mining and size of the each document is different. It is a time consuming process. In this thesis we proposed new web mining method called Web Mining using Divide and Conquer Approach (WMDCA). It consists of four phases: document selection phase (list of documents selected), Preprocessing phase (Divide big size documents, cleaning of each document, combine all sub documents to create XML cube), web mining phase (apply our algorithm to identify patterns), presentation phase (presentation of discovered results). Experiments are conducted on various web documents that are related to one domain. Experimental results of proposed system produce patterns with less time compare with existing methods on web document mining.