Research And Design Of Junior Middle School Education Q&A System Based On Deep Learning

Answering questions occupies a very important position in course education.But the traditional face-to-face approach to answering questions,both for educators and learners,has gradually failed to meet the needs.However,the number of automatic question answering system in junior middle school is relatively small,and there are many deficiencies on the other hand.With the rapid development of smart mobile devices,it has gradually become an indispensable communication tool for people,and even has been popularized in primary and secondary schools.Under such a background,people have been able to accept online learning beyond traditional teaching methods and enhance their knowledge reserves.In the traditional teaching mode,teachers can no longer provide tutoring education for students after they leave the classroom.Therefore,mobile online education such as automatic question answering system has become a good supplementary teaching method.Q&A system is a research hotspot in the field of NLP(Natural Language Processing)and machine learning.It allows users to ask questions directly in natural language,and directly returns concise and accurate answers.It also avoids the return of a lot of useless web links.In recent years,deep learning has made great progress in the field of Natural Language Processing,which makes it possible to apply it in the field of junior high school education.The traditional Q&A system based on traditional retrieval techniques has many problems,such as lack of text semantic description,semantic feature extraction,and poor processing of complex statements.In order to fill in the lack of q&a system of education in junior high school,this paper applies deep learning and word vectors and constructs a question-and-answer system model based on deep neural network.Therefore,in this paper,we first use convolutional neural network and recurrent neural network to build a hybrid deep neural network,so as to better learn deep features in sentences.Secondly,on the basis of the Gated Recurrent Unit,the Bidirectional Gated Recurrent Unit Neural Network and attention mechanism are introduced to improve the weight of the more representative words in the training,so that the model can better learn the semantic matching relationship.Then,we take the data set of junior high school biology as an example to compare and analyze a variety of Q&A system methods,and prove the validity of this model through the experimental results.Finally,we use the Flask framework to build a Web browser based Q&A system Demo,and prove its application in junior high school education.