Ranking user’s comments by use of proposed weighting method

The main challenges which are posed in opinion mining is information retrieval of large volumes of ideas and categorize and classify them for use in related fields. The ranking can help the users to make better choices and manufacturers in order to help improve the quality. As one of the pre-processing techniques in the field of classification, weighting methods have a crucial role in ranking ideas and comments. So, we decided to offer a new weighting method to improve some other similarmethods, especially Dirichlet weighting method. In this paper, the proposed method will be described in detail, and the comparison with the three weighting methods: Dirichlet, Pivoted and Okapi also described. The proposed weighting method has higher accuracy and efficiency in comparison to similar methods. In the following, user comments of online newspapers are ranked and classified by use of proposed method. The purpose is to provide more efficient and more accurate weighting method,therefor the results of ranking will be more reliable and acceptable to users.

کلیدواژه ها:

Opinion mining ، Information retrieval ، ranking comments ، weighting methods ، weighting methods constraints