Question Answering System Analysis Based on Machine Learning

The question answering system is an essential task of natural language processing. In recent years, the deep neural network has made remarkable achievements in many fields. It has reached a level close to human beings in many applications, such as image classification. However, due to the particularity of the question-answering task, the question answering system close to the intelligence level is still a challenge. Firstly, this paper reviews the development of a question answering system, including the Turing Test and traditional question answering systems. Then, the advanced question answering system at present is introduced and analyzed in detail. In addition, we discussed the challenges faced by the current question answering system and puts forward some constructive solutions, including the bottleneck of statistical learning methods, data privacy, and other issues.