DEVELOPING CHATBOT OF ACADEMIC REGULATIONS

Abstract

Most of the students do not read all the contents of the academic regulations, so they often cannot grasp the contents of the organization and training of the major, evaluate the academic results as well as graduation conditions. The main reason is that the academic regulations are too long, students feel bored and take a lot of time to read. Therefore, creating a chatbot that can automatically respond to the content of academic regulations is essential. This thesis will present content related to the theoretical basis as well as implementation methods to create chatbots such as using word processing to process data, applying BoW (Bag of Words) model to build vector for Vietnamese text. Besides, data training and predictive modeling will be performed by Support Vector Machine (SVM) algorithm. The chatbot will understand the question through the similarity of the vector between the input question of user and the question-answer dataset used in the training step. Finally, we applied the above solution to install a chatbot that can answer questions about learning regulations for students.