Identifying Informative Web Content Blocks using Web Page Segmentation

Abstract

Information Extraction has become an important task for discovering useful knowledge or information from the Web. A crawler system, which gathers the information from the Web, is one of the fundamental necessities of Information Extraction. A search engine uses a crawler to crawl and index web pages. Search engine takes into account only the informative content for indexing. In addition to informative content, web pages commonly have blocks that are not the main content blocks and are called the non-informative blocks or noise. Noise is generally illogical with the main content of the page and affects two major parameters of search engines: the precision of search and the size of index In order to improve the performance of information retrieval, cleaning of Web pages becomes critical. The main objective of proposed technique is to eliminate the non-informative content blocks from a Web Page. In the proposed technique, the extraction of informative content blocks and elimination of non informative blocks is based on the idea of Web page Segmentation. Here, a web page is divided into n blocks and the block importance is calculated for each block. The blocks with importance >=threshold are considered as important blocks and the remaining blocks are eliminated as noisy blocks. The proposed approach saves significant space and time