Design And Implementation Of Intelligent Course Question Answering System In MOOC Environment

With the rapid development of social information,more and more people have changed their learning mode from traditional mode to online courses.At the same time,rapid development of Massive Open Online Course(MOOC)has produced a large number of open course platforms.The open courses in China is based on the MOOC platform,and automatic question answering is an important auxiliary teaching mode in the MOOC platform which has been became a hot topic in this mode.Although the question answering system has come of age,but perfect question answering system based on the curriculum knowledge is sparse,and the real-time question answer system has serious problem in real-time performance.Therefore,an intelligent course question answering system for MOOC platform was designed in this dissertation.The system combines web and local Frequently Asked Questions(FAQ)database to search,organize and manage answers.The system calculates the similarity between the user-inputted question and the question in the local FAQ database to find appropriate answer,and if necessary,searches in web and generates the answer using automatic summary extracting technology.The system assists the teacher's teaching,so as to help students get the answers for the related questions of the course in time,guarantee the smooth progress of study,and also promote the improvement of teaching quality and effect.First of all,we researched the method of realizing the intelligent question answering system on the MOOC platform.By constructing the local FAQ database and using Internet search technology,the questions raised by the users are analyzed and the relevant answers are returned,which can make up for the deficiency of online-manager mode for providing answers in current MOOC environment.Secondly,we researched the method of text proofreading for user inputted questions in the process of question analysis.The method use domain dictionaries and N-gram model to rationalize the user's questions,and effectively detect whether the user's question has errors in the word,and furthermore,give the user a good modification suggestion for the wrong input question.Finally,we researched the rational and effective structure of FAQ database and the method of intelligent growth of question and answer contents.The question and answer contents in the FAQ database is increased by using the web search engine and summary processing technology to provide the users with the recommended answers in the FAQ database.The test results show that the text proofreading function and the web-based multi-document automatic generating answer summary function designed in this dissertation achieve the predetermined design goal.The correction rate of the text proofreading function of the question sentence is 85.3%;the accuracy rate of the answer summary based on web multi-document generation is 78.8%;the accuracy rate of the system is 87.3%,the recall rate is 89.3%,and the F value is 88.3%.The test results meet the needs of users.