In this paper, we propose a method for giving answers to users' questions about department information as real-time chatting scheme for fast and accurate search. Firstly, department information is converted and saved as text file, which is analyzed us...
In this paper, we propose a method for giving answers to users' questions about department information as real-time chatting scheme for fast and accurate search. Firstly, department information is converted and saved as text file, which is analyzed using a morphological analysis system. Then TF-IDF weights for the extracted keywords from the morphological analyzer are computed and saved. Similarily, keywords are extracted from users' questions, then high weighted answers are displayed after matching. Through those steps, appropriated answers can be given to users. Experimental results show that the proposed method achieved 70% accuracy in getting correct answers. We also developed a smart phone web app to show the utility of the proposed method.