Many research organizations and companies are conducting research on intelligent Q&A, but their technical level has to be further improved.
The current research on open-domain Q&A systems mainly focuses on single, isolated, factual questions, and is limited in the accuracy of the answers as well as the complexity of the questions by the complexity of the existing computational models and the degree of perfection of the knowledge base.
Watson is unable to provide satisfactory answers when it comes to semantic comprehension, complex logical reasoning, and chapter-level linguistic analysis; AI2 has so far only reached the biological level of a fourth-grade student.
With the support of deep learning, knowledge mapping and other technologies, there is a possibility of a huge breakthrough in algorithms, knowledge and data for Q&A systems.
It is reasonable to believe that in the near future, computers will realize highly intelligent automated Q&A, and the technologies developed will bring about technological innovations in artificial intelligence, which will have a far-reaching impact on the information society.