Context based user ranking in forums for expert finding using WordNet dictionary and social network analysis

Currently, online forums have become one of the most popular collaborative tools on the Internet where people are free to express their opinions. Forums supply facilities for knowledge management in which, their members can share their knowledge with each other. In this regard, The main problem regarding to the knowledge sharing on forums is the extensive amount of data on them without any mechanism to determine their validity. So, for knowledge seekers, knowing the expertise level of each member in a specific context is important in order to find valid answers. In this research, a novel algorithm is proposed to determine people's expertise level based on the context. AskMe forum is chosen for the evaluation process of the proposed method and its data has been processed in several stages. First of all, a special crawling program is developed to gather data from AskMe forum. Then, raw data is extracted, transformed, and loaded into a designed database using SQL server integration services. Afterwards, people's expertise level for specified context is calculated by applying the proposed method on the processed data. Finally, evaluation tests are applied in order to calculate the accuracy of the proposed method and compare it with other methods.