Term impact-based web page ranking
|Term impact-based web page ranking|
|Author(s)||Al-Akashi F.H., Inkpen D.|
|Published in||ACM International Conference Proceeding Series|
|Keyword(s)||indexing, query expansion, searching, term impact, vector space model, Web retrieval, Wikipedia anchors (Extra: Indexing (of information), Search engines, Semantic Web, Semantics, Vector spaces, Query expansion, searching, term impact, Vector space models, Web retrieval, Wikipedia, Websites)|
|Article||BASE, CiteSeerX, Google Scholar|
|Web||Ask, Bing, Google (PDF), Yahoo!|
|Download and mirrors|
|Local copy||Not available|
|Remote mirror(s)||Not available|
|Export and share|
|BibTeX, CSV, RDF, JSON|
|Browse properties · List of conference papers|
Indexing Web pages based on content is a crucial step in a modern search engine. A variety of methods and approaches exist to support web page rankings. In this paper, we describe a new approach for obtaining measures for Web page ranking. Unlike other recent approaches, it exploits the meta-terms extracted from the titles and urls for indexing the contents of web documents. We use the term impact to correlate each meta-term with document's content, rather than term frequency and other similar techniques. Our approach also uses the structural knowledge available in Wikipedia for making better expansion and formulation for the queries. Evaluation with automatic metrics provided by TREC reveals that our approach is effective for building the index and for retrieval. We present retrieval results from the ClueWeb collection, for a set of test queries, for two tasks: for an adhoc retrieval task and for a diversity task (which aims at retrieving relevant pages that cover different aspects of the queries).
- This section requires expansion. Please, help!
Probably, this publication is cited by others, but there are no articles available for them in WikiPapers.