A Research On Domain Expert Finding In The Social Network

In recent years, people are increasingly inseparable from the social network, Almost every ordinary people and experts have established their own account on the social network, used to sharing and receiving information. Social networks allow users easy to obtain large amounts of information, and it is convenient to connect almost anyone online. When encounting problems, more and more users want to find the solution through the social network. there are usually two ways to solve problems online:Firstly, we can obtain the solution through the retrieval of information on the social network, such as Baidu zhidao, zhihu QA,Facebook and Twitter. These network platform has accumulated a large number of useful information. Secondly, we can consult an expert. But the users of social network may be more than several hundrde million, so it is hard for us to find the expert in a given domain. It is an urgent problem to be solved. The application of expert finding provides a new way to solve this problem.Expert finding, as the name suggests is a method that could be used to identify experts, so we can solve the problem by consulting the right experts. For example, enterprises usually encounter technical problems which can not be resolved by themselves, then whom to consult for help is very important. The expert finding method is widely used in scientific research. Several good approachs have been carried out. These methods make use of the academic circle, email correspondence, employee files to achieve the experts.In view of the above reasons, this paper will apply expert finding method to the a social network named weibo. we make use of local information and online behavior to find the experts in a given domain. To facilitate social network users find the corrcect expert for solution.Many traditional research find experts through link analysis method. But when muti topics exist in the social network,this method does not work well.And in the social network of weibo, users concerned about various topics, which almost involving all aspects of life. In order to solve this problem, in this paper, we make use of topic model for constructing the domain dictionary. So we can limit the target user’s topic through the domain dictionary.In orde to find and evaluate the expert we take links between the users and the personal information into consideration. At last the expert finding model identify and return the highest ranked experts. Based on the platform of sina micro-blog,we carry out the expert finding model to find users who are experts in the domain of data mining.