Research And Implementation Of Intelligent Question Answering System

With the emergence of the online education mode,many teaching institutions began to lecture online,more and more students are also participating in the online learning team.Unfortunately,in the process of online teaching,teachers and students can not discuss and communicate in time because of the separation.It causes difficult problems of students can not to be solved.It also hinders students' study to a great extent and affects students' learning.The question answering system plays an important role in network education,and it is more important to establish an intelligent answering system which can answer students' questions efficiently and quickly as the students' assistant.The main goal of this paper is to design and develop an intelligent question answering system based on the PAR method teaching website,and to complete the answer to the related questions in the field of program design.First of all,it analyzes the characteristics of the question and answer system generated in the network education,introduces the research status of the question and answer system at home and abroad.Second,an overview of the key technologies used in this question and answer system is presented,a Markov prediction model is proposed,and parameters are set through experiments.Third,the overall design of the answering system and the design of each functional module are completed.Finally,the realization of related key technologies of the system is presented.The functional modules of the question and answer system are displayed.The accuracy of the system's question and answer is verified by the application testing of the intelligent answering system.The main innovation work of this paper lies in:(1)This article developed a smart question answering system based on natural language processing technology.The system has successfully completed the processing of Chinese natural language questions.It uses the forward maximum matching algorithm for Chinese word segmentation and uses a keyword-based sentence similarity calculation method and a multi-information sentence similarity calculation method based on a synonym word forest-based semantic similarity calculation method to calculate the similarity of two sentences.This method not only considers the degree of similarity between sentences,but also considers the semantic similarity of sentences,so that the system can more accurately understand the needs of students,thus giving correct answers to questions.The natural language processing technology is introduced into the answering system,which greatly improves the accuracy of the system answering questions.This is one of the innovation points.(2)This paper applies the hidden Markov model to the intelligent question answering system for the first time.First of all,this paper collects relevant questions and answers in the field of program design and classifies the knowledge into these questions to obtain a data set.Then using a large number of question sequences to train and obtain a hidden Markov model,which means to determine the probability of association between the knowledge points;Finally,according to the students' current question and the known Hidden Markov Model,the migration of the internal knowledge points is predicted,the students' most likely next question is obtained,and the function of the system to the student's intelligence recommendation problem is finally achieved.Constructing a Markoff question prediction model in the intelligent question answering system greatly improves the efficiency of answering questions.This is the second part of the innovation point.