This paper discusses the use of Wikipedia for building semantic ontologies todo Query Expansion (QE) in order to improve the search results of searchengines. In this technique, selecting related Wikipedia concepts becomesimportant. We propose the use of network properties (degree, closeness, andpageRank) to build an ontology graph of user query concepts which is deriveddirectly from Wikipedia structures. The resulting expansion system is calledWiki-MetaSemantik. We tested this system against other online thesauruses andontology based QE in both individual and meta-search engines setups. Despitethat our system has to build a Wikipedia ontology graph in order to do itswork, the technique turns out to work very fast (1:281) compared to anotherontology QE baseline (Wikipedia Persian ontology QE). It has thus the potentialto be utilized online. Furthermore, it shows significant improvement inaccuracy. Wiki-MetaSemantik also shows better performance in a meta-searchengine (MSE) set up rather than in an individual search engine set up.
展开▼