Design And Implementation Of Campus Information Question Answering System Based On IPV6

With the rapid advancement of science and technology,the application of artificial intelligence into the whole scene of education is increasing day by day,and the application of "artificial intelligence + education" products has a lot of room for improvement.The campus information question answering system is one of the applications,which is used to serve teachers and students to quickly capture the answers they need in a large amount of information and meet their daily information consultation needs.Meanwhile,IPV6 technology is the trend,and this paper implements a campus information question answering system based on ipv6,which is in line with the development direction of the times and has wide research value and application prospects.This paper will research and develop a campus information question answering system based on IPV6,and make an in-depth study of the key technologies used in it,the main work is as follows:(1)Automatically constructing a knowledge base for campus information.In this paper,we use python language combined with crawler technology to crawl campus data information and obtain the source knowledge of campus information.Then uses the "BERT + Uni LM" approach to construct the Seq2 Seq model to achieve end-to-end construction of question answer pairs from the source knowledge,and finally obtains a knowledge base of questions and answers on campus domain FAQs.Through experiments,this paper proves that the method of automatically building campus knowledge base is efficient and feasible in practical applications with high accuracy.(2)Text similarity matching method.A text similarity matching method based on the combination of Siamese network and Char-Word vectors is proposed.The BERT +Wo BERT model with the combination of Char-Word vectors is used to solve the problem that the traditional model can hardly focus on the semantic-grammatical information of Chinese texts.Then exploring the impact of multiple fusion methods and dimensional noise reduction on similarity matching results through Siamese network and PCA algorithms.The binary classification is then performed by Softmax,which ultimately achieved better similarity matching results on the dataset.(3)Design and implement a campus information question answering system based on IPV6.Based on the findings of this paper,existing development tools and frameworks are used to apply the system.Use the Django framework to open services to achieve access under IPV6,and the front-end interface?user login and questioning?back-end matching model invocation and database addition?deletion and checking functions are developed.The test and analysis results show that based on the self-built campus question answering knowledge base,the campus question answering system developed in this paper runs stably and achieves satisfactory results in terms of response rate and answer accuracy.