Design And Implementation Of An Intelligent Question Answering System For Small Loan Consulting Business

In recent years,with the development of global network communication and information technology,many online businesses are gradually replacing the traditional offline businesses.The microfinance consulting business of loan institutions also faces the same business transformation.The traditional customer service system is facing the problems of continuous increase of labor cost and low efficiency of customer service.Therefore,the design of an online intelligent question answering system can replace the artificial customer service to accurately and effectively answer a large number of repeated questions raised by users.Based on the demand of microfinance consulting business,this paper studies and develops an intelligent question answering system,which mainly includes:1.Using the Django Web framework technology to build a user question answering interface and administrator management interface of the intelligent question answering system.Users can enter questions and get answers to the questions,and administrators can add,delete,modify,and maintain the data in the knowledge base.2.The server searches for answers to the questions entered by the user.If the search fails,it can be identified as gossip,calling a third-party interface to continue to answer the questions for the user;if the retrieval is successful,the questions entered by the user and the knowledge base retrieved the question and answer pairs match the answers,find the best answer and return it to the user,and at the same time do emotional recognition on the user's questions.For users with negative emotions,they automatically send emails to warn the intervention of manual customer service to improve user satisfaction.The question and answer matching module first uses answer retrieval technology,and then uses a vector space model(Vector Space Model,VSM)and Word2 vec mixed question similarity algorithm to find the best answer.Gradient lifting classification algorithm of iterative decision tree voting election mechanism.This paper describes the above work from the aspects of technology and algorithm introduction,requirements analysis,design,system implementation and testing.