Recommending Answerers for Stack Overflow with LDA Model
Bin Shao, Jiafei Yan · 2017
Stack Overflow is the largest Community-based Question Answering (CQA) site for software developers. Its popularity is mainly attributed to the timely answers provided by a great number of developers. When having problems in learning and using new technologies, developers resort to Stack Overflow for help. However, it is difficult to recommend questions to the potential answerers due to the huge numbers of questions. In order to improve the accuracy of question recommending, we need to find out the members who are interested in the fields related to the questions and match the ability of developers with the difficulty of questions. To do so, we need to pay close attention to the behavior of developers. This paper presents a model with two prediction approaches, namely, the traditional feature-based approach and LDA (Latent Dirichlet Allocation) based approach. When a new question arrives, this model will use LDA method to label the question and indicate the proper category to which the question belongs according to latent semantic feature and content feature. Then, with the traditional features of the question and the asker information, the model will recommend the appropriate developers to answer this new question.