Font Size: a A A

Design And Implementation Of An Enterprise Customer Service System Based On Online Chat Robot

Posted on:2021-03-30Degree:MasterType:Thesis
Country:ChinaCandidate:Z H ZhouFull Text:PDF
GTID:2518306104496064Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Nowadays,the competition among enterprises is not only reflected in product quality and price,but also in service.A convenient,efficient and functional enterprise customer service system can be an important means in order to improve the customer satisfaction rate and enhance the market competitiveness.A customer service chat function module enterprise is embedded in the existing customer service system or portal website.Through the development and construction of customer service module,the online communication and instant information consultation platform can be provided for visitors and customer service.Online intelligent chat robot can solve a large number of similar or repeated questions raised by users with the help of natural language,query and retrieve the knowledge base,and then answer with more accurate and simple results,so as to reduce the human cost of customer service and improve the efficiency of customer service.Therefore,it has great practical value for the research and implementation of enterprise customer service system based on online chat robot.Based on the large-scale work order processing business,the design and implementation of enterprise customer service system based on online chat robot mainly includes the architecture and development of enterprise customer service system,as well as the design of key algorithms of online chat robot.The main work is putting forward the software architecture composed of front-end display layer,business logic layer,data processing layer and data storage and access layer,which is targetting the enterprise customer service system based on chat robot;The k-means algorithm combined with naive Bayes algorithm is used to classify text,and the semantic analysis similarity algorithm based on Word2 vec is improved,which is weighted and combined with the structural analysis similarity algorithm based on word order and word shape similarity to match the similarity of questions.The development of enterprise customer service system adopts the B/S mode,using the relatively advanced development framework or technology,such as Spring Boot framework,Web Socket technology,non-relational database Redis,etc.The application of new technology improves the stability and scalability of the system,and ensures the service life of the system in the current Internet environment.After testing,the accuracy of the chat robot module in the restricted domain is 78.5%,which reaches the industry average level.After the first version of the system is put into trial operation,nearly 50% of the customer service labor cost is saved,the customer service processing efficiency is accelerated,the customer's praise is gained,and the system has high commercial value.
Keywords/Search Tags:Customer Service System, Chat Robot, Naive Bayes Algorithm, Word2Vec Model
PDF Full Text Request
Related items