User’s intent detection in education

• Main Author: Van, Thi Hoai Thuong
• Other Authors: Tran, Thi Oanh
• Format: Final Year Project
• Language: Vietnamese
• Published: 2020
• Subjects: User Intent Detection; Educational chat hot; SVM; RCNN
• Online Access: http://repository.vnu.edu.vn/handle/VNU_123/98148
• Institution: Vietnam National University, Hanoi

The transition from traditional consulting and advertising to digital consulting and advertising in recent years depends heavily on scientific achievements, specifically, achievements in machine learning, natural language processing and artificial intelligence. In addition to these conversions, the automatic question and answer system - chat bot system is increasingly being popularly used and tends to rise not only to attract businesses but also to attract educational institutions and universities. In addition, the development of the internet and social network makes people tend to search for information and get information online more than traditional approaches. With a useful virtual assistant like chat bot system, improving the user experience or saving resources for the organization is something that has reached a higher level. The idea of research is to identify the intent in the first step of the process of building an educational chat bot system. With the statistical traditional approach, use the SVM (support vector machine) the accuracy is 93.93 % (TF-IDF & 1:2 Grams) and thee model using deep learning perform 92.34% (RCNN).