Predicting Best Answerers for New Questions: An Approach Leveraging Distributed Representations of Words in Community Question Answering

Community Question Answering (CQA) sites are becoming an increasingly important source of information where users can share knowledge on various topics. Although these sites provide opportunities for users to seek for help or provide answers, they also bring new challenges. One of the challenges is most new questions posted everyday cannot be routed to the appropriate users who can answer them in CQA. That is to say, experts cannot receive questions that match their expertise. Therefore new questions cannot be answered in time. In this paper, we propose an approach which based on distributed representations of words to predict the best answerer for a new question on CQA sites. Our approach considers both user activity and user authority. The user activity and user authority are based on the previous questions answered by the user. We have applied our model on the dataset downloaded from StackOverflow, one of the biggest CQA sites. The results show that our approach performs better than the TF-IDF and Language Model based methods.