A NOVEL APPROACH FOR DIGITAL LIBRARY SEARCH RESULTS REPRESENTATION USING CLUSTERING AND RANKING

It is a vital task to rank the digital documents in digital libraries due to an exponential growth of digital information and need of users. Ranking mechanism plays an important role in digital libraries as it enables the user to find the desired document efficiently. Various ranking algorithms have been proposed based on different measures like number of citations to a research paper, content of paper, impact factor of publication venue, published year of the paper, bookmarks etc. But, these existing ranking algorithms sometimes provide irrelevant results due to certain shortcomings, which indicate a scope for further improvement in ranking mechanisms. In this paper, an optimized ranking algorithm is proposed that carries out static as well as dynamic ranking to rank the documents in digital libraries. The proposed algorithm considers the citations of the paper, bookmarks of the paper, user’s feedback and clustering process for ranking. An optimized approach is being proposed which provides sorted search result list in cluster form against the user’s query.