An intelligent question-answering system for course learning based on knowledge graph

Abstract

The continuous development and digital transformation of the education field has caused educational institutions and students to face huge challenges in information acquisition and learning resource management. The demands of modern learners cannot be fully satisfied by conventional teaching techniques, and knowledge graphs and Natural Language Processing (NLP) technologies have become a potential remedy. Students can benefit from a personalized learning experience and enhanced comprehension and mastery of material through the knowledge graph-based intelligent question and answering system for course learning.

This research is dedicated to developing a knowledge graph-based intelligent question and answer system for courses learning, allowing users to view the knowledge graph and ask questions. The system will adopt deep learning methods, selecting the Pytorch deep learning framework and Transformers natural language processing deep learning model. Specifically, it is the Joint BERT model, which is a BERT model based on user intent recognition and slot filling. After analyzing user input, corresponding reply statements are generated by matching the question and answer template and querying the Neo4j graph database. Ultimately, intelligent question and answer systems create effective and engaging learning interactions for users.