Research On The Design Of Online Personalized Question Answering Based On LSTM Neural Network

On April 13,2018,China's Ministry of Education issued "Education Informatization 2.0 Action Plan",emphasizing the need to construct a "internet plus" talent training model and develop a new model of education services based on the Internet.In August of the same year,the 2018 Kwun Tong Declaration on Intelligent Education was issued.Ten domestic experts in the field of information education also proposed to explore the deep integration of artificial intelligence and education,innovate education and teaching modes,and build a new system of intelligent education.The application of new artificial intelligence technology to strengthen the intelligence of educational services has gradually become the focus of attention.With the development of new artificial intelligence technology,the automatic question answering system,as a support service for online courses,has attracted more and more attention from experts and scholars in view of the current situation that learners' questions cannot be answered in a timely and effective manner in the online course learning process.Based on the research and analysis of distance education theory and personalized learning literature,increasing dialogue and interaction can effectively improve online course learning and personalized learning services.This research attempts to help learners complete personalized learning of online courses through the automatic question-answering system as a medium,and explores the design and research of automatic question-answering system based on depth learning technology.This article is carried out through literature research,design-based research,observation and interviews.Through literature research,the research status and implementation of automatic question answering system at home and abroad are analyzed,and the design idea of this research is established.According to the design idea,further literature research and analysis are carried out on the deep learning neural network model,and the LSTM neural network model suitable for this research is selected.Through organizing learner learning tests in the early stage,the learning problems existing in the learning process are collected,and based on this,a data set suitable for the training of the network model in this study is constructed.Then,a neural network model is constructed and three rounds of iterative optimization design research on the recognition accuracy of the model are carried out.Continuous adjustment enables the recognition accuracy of the neural network model to reach a higher value,supporting the development of project experimental tests.Then,through the program design,the recognition and answer of the learner's input questions are realized through the similarity detection of the input questions.After the automatic question-and-answer is realized,the learners are organized to carry out online course project experiment tests.In this process,the experimental data of the learners are observed and recorded through observation and interview methods.Through the research and analysis of the experimental test data,the influence of personalized question-and-answer of the learners' online course on the learning effect and the depth of thinking of the course content is explored.