Wikipedia based Query Expansion for Searching in Norwegian

Abstract

With the rapid increase of information available on the Internet, searching is becoming a daily operation. This thesis address the issue that arises when there is a difference between what the users are typing in the search engine, and what they actually are interested in. The method presented in this thesis is a Query expansion method using Wikipedia as a knowledge-base, with the goal of reducing the gap between what the users are formulating, and the actual information need behind the search. The method is based on analysis of hyperlinks between articles in Wikipedia. The research contribution of this thesis is a new method for query expansion designed for Norwegian language. The proposed method uses the knowledge-base to suggest words for the users that are relevant to the original query. The evaluation of the new method shows a marginal increase in quality, but highlights the need for a search engine designed for Norwegian language. The thesis concludes that a search option for Norwegian language will produce better results, and the system for suggestion of words might help users move from a broad search strategy to a more precise search strategy.