Focused Crawling aims to search the WWW in relevance to the user topic of interest, tends to provide more number of relevant documents in first top results, leading to the need of maintaining the index with more number of related web pages; using the relative measure of relevancy between the documents. This paper provides a novel technique for extracting the back-links of a web page and to evaluate their context score; that helps to update the document index with more number of relevant documents.