A data mining approach to predict users' Next question in QA system

摘要

In a Question Answering system, the user submits a question and waits for the answer as the response. If the system is capable of predicting the user's future interest as the next question, its performance will improve greatly. This paper predicts users' future questions based on the current interaction records of the user with the system. Their current interactions with the system show what they are interested in. These interaction records are maintained in the form of Questions log from which the user sessions are extracted. Based on the user sessions, the system predicts the next question for which the user may become interested in near future. A sample Questions Log is selected for the purpose of performing experiments. The model of Association rule mining is applied to predict the future question of the user.