Intelligent Interpretation of World Wide Web Queries

Abstract

The problem of finding information on the Internet or on intranets is becoming increasingly more difficult, because of its explosive growth and lack of structure. This paper describes IWWWQ, a system that improves the ranking of a search engine's results by incorporating other features of a page, including the page type, content of a page, page formatting, and the relationship between pages. An empirical evaluation shows that incorporating these features improves the ranking of the results returned by a popular search engine. Furthermore, the evaluation showed that a combination of features is necessary to obtain this improvement. Finally, it shows that the algorithms used are relative insensitive to specic parameter settings.