Design And Implementation Of Question Answering System For Assisting Online Learning

The rapid development of the Internet has promoted the process of information reform in the field of education.Through various learning platforms,learners can easily access rich digital resources,but the phenomenon of "knowledge lost" is becoming more and more common.Not only users can not locate the required knowledge content,but also the platform often lacks the means to fully integrate and utilize the education big data.Compared with the retrieval ability of search engine,the auxiliary question answering system is a powerful method to improve the accuracy of knowledge acquisition.Aiming at the common scene of online learning,from the perspective of solving knowledge retrieval and enhancing the interaction experience of learners,combined with the basic model of natural language processing,this paper puts forward the auxiliary question answering model for online learning,and designs and implements the corresponding question answering system based on the model.The main work of this dissertation is as follows:(1)This paper analyzes the types of Knowledge Q&A in online learning scene,extracts the related Q&A pairs in education field through network data and open corpus,uses many natural language processing technologies such as word segmentation and word embedding to identify entities and relationships to form triples,and uses entity alignment and disambiguation methods to integrate multi-source knowledge resources to build knowledge represented by RDF Library.(2)The model structure and training scheme are designed to verify the effectiveness of two application scenarios:knowledge query and teaching task guidance.For the knowledge query-oriented Q&A model,the shallow neural network based on seq2seq structure is adopted,and the results of semantic analysis and search sorting are integrated to return the answers.As for the question answering model for teaching task,we use the rule dominated finite state machine to define the use cases and boundary constraints manually,which provides more application scenarios with scalable and applicable examples.(3)According to the existing models and strategies,based on the consideration of model embedding and iteration,and high concurrency,we design and implement the whole architecture of online learning oriented auxiliary question answering system,as well as the details of front-end,back-end and database,complete the knowledge query,teaching task guidance,knowledge base management and other functional modules,and finally test the function and performance of the system,check its stability.