In recent years, community question answering system (CQA) are becoming popular. But with the expansion of the scale of the questions, the proportion of questions that got answered gradually reduced, and the quality of the answers cannot be guaranteed. In order to increase the answering probability of the questions in Q&A system, and enhance the credibility of answer, we put forward Social question answering System based on the social relationship similarity measure. Then we raise a method to find suitable respondents who are willing to answer and are familiar with related field. Experimental results show that the method of this paper can get satisfactory answers faster compared with the traditional Q&A System on the subjectivity or real-time problem sets.