Enhancing Wikipedia search results using Text Mining

Wikipedia is an online encyclopedia which contains millions of articles related to different subject domains. Wikipedia also has a search page itself to display the links corresponding to Wikipedia articles for a given user query input. This search result page displays the search results according to the relevance order, without any content based grouping. This paper presents an experimental deduction of a search result clustering methodology to group the links, returned by the search result page for a particular keyword, based on the contents of the HTML documents, represented by the links and label these resulted groups meaningfully. The proposed methodology is based on the concepts and theories of Text Mining. Grouping of search results makes easy and efficient for the user in finding the desired Wikipedia document. It is also possible to view the different applications and usages of a given keyword very quickly. This work identifies the best clustering algorithm for document clustering and investigates the ways to determine optimum number of clusters to have a better grouping and label the groups. We evaluate our proposed method by conducting several experiments and the results indicate that our method has a higher precision and recall.