Semantic networks are some kinds of networks that model the conceptual relations among words. WordNet is a lexical database and is an instance of such semantic networks that assort the words on the basis of their semantic similarities. Nowadays, availability of such lexicon-semantic networks seems to be serious in many natural language processing applications and information retrieval tasks. Hence, construction and extension of WordNet, as one of the most significant knowledge resources, is considered in the plenty of languages in recent years.
In this thesis, our endeavor is to increase the accuracy of automatic approaches proposed for Persian WordNet construction. On the other hand, we focus on the improvement of Persian words coverage in the current version of Persian WordNet. This purpose will be done through increment the number of synsets and their links to Persian words. In this thesis, we will construct the Persian WordNet from two points of view. First, we focus on the previous related works in this field and try to improve them. Second, we endeavor to present a new approach for Persian WordNet construction by exploiting other language resources.