Routing Questions to Experts in Community QuestionAnswering Sites
Anurag Modi, Manish Singh · 2018
In CQA sites, user can ask questions, which are answered by other users having expertise in that domain. These sites allow the knowledge exchange between users in terms of question and answers.This knowledge base of the question-answers which is getting created also serve as a reference to the other users who in later future may ran into similar problem. But due to large volume of these questions getting asked every hour, all questions do not end up getting answers. As a result, it can be useful to route a question to the potential answerer. A lot of prior work done in the field of the question routing, focuses on the static snapshot of data, whereas these sites are dynamic in nature i.e. users constantly join the site and user topic of interest might get change over time. Expertise �nding strategy used so far only considers the static nature of these datasets and do not model the change in the users behaviour. Some work [5] has been done on �nding availability of user at a particular point of time. In this paper, we propose a recommendation model that does tag network and temporal analysis to find the potential users who will answer the question. Our experiments over a large real-world datasets from Stack Exchange shows effectiveness of our approach over several baseline models. Our results shows that considering temporal information and community valuation of answer to compute user expertise provides improvement over the baseline models used so far for expertise recommendation, which shows the effectiveness of our proposed model.