YONA: An Intelligence Question-Answering System Based on Artificial Intelligence

The purpose of this paper is designing an intelligent question-answering system. In the context of education informatization, university students tend to lack guidance in their studies. This paper proposes an intelligent Q&A system for university teaching that integrates intelligent dialogue and machine learning technologies. The proposed system includes intelligent Q&A module, the tutor recommend & match module, tutor Q&A system module, Q&A forum Module and user compasrison report module. The system utilizes BERT model, collaborative filtering recommendation algorithm, multi-threaded programming, TF-IDF algorithm and K-Means algorithm to implement the modules in the system. The proposed system deployed with Aliyun cloud server, CDN, load balancing, client testing, and online optimization. The pilot of this system demonstrated excellent performance: an average of 4 queries per person, with an average query duration of 7 minutes. The success rate of problem resolution reached 85%, and the average response time was 30 seconds. This system will promote the health of the education industry and enable intelligent learning diagnosis. The authors hope to contribute to improving the quality of life of college students and society by designing an intelligent question and answer system.