Social networks usage increased significantly in recent decades. The range of data in these networks are high because usage of these kinds of networks are common among people, hence, categorizing these data could be really useful. Nowadays, some of these networks use a method for categorizing data which is called Hashtag. Using Hashtag, users can find messages and posts in a specific social network and find their desired subject regardless of the meaning and semantic of the Hashtag. Defining a new approach for categorizing Hashtags is a new and practical approach for categorizing data in social networks. In this thesis we provide a new approach for categorizing Hashtags in the social network of Twitter In our proposed approach, we use hierarchical subject clustering of comprehensive library of congress. We index the text information of our sources, on the other hand, keywords related to messages containing specific desired Hashtags, using semantic and text mining algorithms are extracted and indexed. After diagnosis and isolation of hashtag creator, using the scoring algorithm which is proposed in this thesis, the nearest categories from the inquiry could be extracted, then through another scoring algorithm, shared data between the extracted data from specific Hashtag could be found and then pointed as a best category for the desired Hashtag. Thus, we will be able to provide a service in which user can get the most related categories by entering the Hashtag and the relevant messages. The ability of using upper and downer branches of a certain category will be provided as well. Therefore, semantic Hashtag is standard categories with closest meaning to Hashtags created by ordinary users of social networks which allows access to reliable sources for users Keywords: Hashtag, Semantic algorithms, categorizing, social networks
