Improving Efficiency of the Focused Web Crawler by Link Score Calculation

Abstract

Abstract
The overwhelming size of the Web enables continuous support and recurring update of Web based information retrieval systems. The Crawler can narrow down the entire process to allow the hyperlinks in Web pages to download a part view of the Web. Meanwhile few systems absolutely depend on crawlers that exclusively crawls the web, and extract the topic specific collections. A focused crawler has the tendency to accumulate particular topics and target to gather relevant segments but not to consume resources on irrelevant material. The sole objective of the focused crawler is to fetch the maximal set of relevant and quality pages. In the proposed approach, the classifier task is to categorize the unvisited URL based on visited URLs attribute score and eliminating the unvisited URLs which are not relevant to the specific domain.

Keywords: crawlers, information retrieval, web mining, link score, robot