Research On Intelligent Question Answer Technoogy Based On Deep Learning

With the rapid development of Internet technology,network information data has grown exponentially.In the era of information big data,in the face of massive information,how to quickly and accurately obtain the required information is an urgent need of users.Compared with the search engine,the intelligent question answering system can directly return the answer information required by the user according to the natural language question submitted by the user,and reduce the time cost of the user to obtain the information.Deep learning technology is changing with each passing day,and it has been widely used in the field of natural language processing.As an important application form of natural language processing,the intelligent question and answer processing system has gradually attracted attention.This paper first analyzes the deep learning techniques used.For the standard gating cycle unit,the semantic vector representation is not able to completely learn the meaning of the sentence.The bidirectional gating cycle unit is semantically encoded to obtain a complete semantic understanding vector.According to the thought of attention,construct a multi-step attention mechanism at the sentence level to improve the accuracy of the neural network model in obtaining the text information vector related to the answer.Then the intelligent question and answer neural network model is constructed by combining Bi-GRU and multi-step attention mechanism,and the accuracy of the model to generate the answer after extracting the information text vector according to the problem is verified by experiments.Finally,this paper analyzes the function and architecture of the intelligent question answering system,and separates the model training and question and answer services to ensure the usability and stability of the question and answer system.