Improving the Relevance of Search Engine Results by Using Semantic Information from Wikipedia

Depending on the user's intention, the queries processed by a search engine can be classified in transactional, informational and navigational [1]. In order to meet the three types of searches, at this moment search engines basically use algorithmic analysis of the links between pages improved by a factor that depends on the number of occurrences of the keywords in the query and the order of these words on each web page returned as a result. For transactional and informational queries, the relevance of the results returned by the search engine may be improved by using semantic information about the query concepts when computing the order of the results presented to the user. Wikipedia is a huge thesaurus which has the advantage of already being multi-lingual and semi-structured, presenting a dense structure of internal links that can be used to extract various types of information. This paper proposes a method to extract semantic relations between concepts considered as the names of the articles from Wikipedia, and then use these relations to determine the rank of the results returned by a search engine for a given query.