Autonomous Question and Answer System Based on ChatGPT With Large Language Model

Chat-GPT has become very popular in recent years. But there is a problem. Chat-GPT does not ask questions to the users. Therefore, Chat-GPT looks like a machine, not a human. However, users sometimes do not want a single answer. They want real things like food, cars, products, etc. Therefore, our system will ask users questions several times until they get what they really want. In this project, we not only resort to Chat-GPT3.5 to find questions, but also resort to traditional programming skills or databases to solve these problems. OpenAI's Chat-GPT3.5 will play the main role in this project. Furthermore, Java and Spring-Boot will be used in this project. These are mature frameworks for enterprise systems(de Oliveira, C. E., Turnquist, G. L., & Antonov, A., 2018). Finally, the MySQL database is used in this project. It provides comprehensive and reliable SQL database services. The data stored in MySQL instances can generate very large data sets(bin Uzayr, S., 2022). Python3 and some machine learning frameworks such as NumPy, Pandas, TensorFlow and PyTorch are used in this project to analyze user behavior(Liu, Y. H., 2020). In the future, the dataset can be integrated into webtext2 as a basic data element to connect the model training.