Research And Implementation Of Socialized Question Answering System For Closed Complex Domain Business Consulting

In closed and complex domains such as universities,operators,and large group companies,a large number of consultations from outside and inside the domain are faced every day,at present,common methods,such as setting up special customer service personnel or publishing information through portal websites,on the one hand,which increase the operation cost of domain,on the other hand,it is not easy for consultants to get answers.This article provides another idea:to implement a social question answering system similar to Zhihu,and solve this problem through knowledge sharing among users.This paper analyzes the current situation of closed complex domain business consulting,designs and implements a socialized question answering system for business consultation in closed complex domain,and solves the problems of high operating costs and low efficiency for consultants to obtain answers in closed complex domains.And analyzes the key and difficult problems in the system,including similar question matching,speech content limitation and answer quality ranking,and gives the corresponding solutions.The main work of this paper is as follows:1.The key problems in the implementation of the system are analyzed,and the corresponding solutions are given.Firstly,for the similar question matching problem,an improved text similarity calculation method based on Sentence-BERT model is proposed,and the top 5 questions in the system that are highly similar to the questioner's question text and their associated answers are returned to the questioner.Then,for the problem of speech content limitation,a sensitive thesaurus expansion method based on Word2vec is proposed,and then based on the sensitive thesaurus,the sensitive words in the speech content of system users are identified by multi-mode string matching method.Finally for the answer quality ranking problem,based on Wilson scoring algorithm,the quality score of the answer is obtained according to the support number and opposition number of the answer,and sort the answers according to the score,and the sorted answer is returned to the user.2.This paper analyzes,designs and implements a socialized question answering system for business consultation in closed complex domain.Firstly,the user roles of the system are divided from the perspective of users.Secondly,analyze the functional requirements,including the rationality and necessity of the functional requirements of tourists,questioners and administrators.Next,determine the logical architecture of the system and design the module structure of the function.Then carry out the database design of the system,including entity relationship analysis and conceptual design.Finally,a socialized question answering system for business consultation in closed complex domain is implemented to solve the problems of high operation cost and low efficiency of consultants in obtaining answers in closed complex domains.