Research And Implementation Of Intelligent Dialogue System Based On Micro-service

With the development of 5G and the increase of AI application scenarios,it will have a huge impact on people's work and lifestyle.Enterprises should seize this opportunity to create better products and provide better services.In recent years,due to the rise of mobile intelligence,conversational services have become a hot research direction.Intelligent assistants will play an important role in this change.Users can interact with computers through language,which will greatly improve the efficiency of human-computer interaction.Question consultation and process management can be completed through dialogue,which will greatly improve the user experience.In commercial applications,the purpose of enterprises is to provide services to users in specific scenarios,so we only need to design a question-and-answer robot to reply to users' inquiries,and a task-processing robot to help users handle business.This paper designs and implements an intelligent assistant based on the microservice architecture in detail,which is a user-oriented dialogue system and a customer service-oriented hybrid intelligent console(HIC).The dialogue system meets the needs of users to answer questions and process processes in the car rental scenario,and the hybrid intelligent console meets the customer service's needs for users to transfer to labor,complaints,and reports.The development tasks of the intelligent assistant include: front-end page,dialogue management,data management,report management,customer service management,business process,NLU service,etc.This paper mainly expounds the design and implementation of the dialogue management module,data management module,customer service management module and report management module in the intelligent assistant.This system uses Python to write server-side code,uses asyncio asynchronous framework and Websocket to complete front-end and back-end data communication,uses SQL Server database to complete data storage services,uses Redis database to complete data caching services,and uses Rabbit MQ message middleware to perform communication between different servers.Message synchronization,use F5 and Apache to achieve load balancing,use IIS to deploy front-end files and configure routing forwarding,and finally use Docker containers to deploy applications in a Linux environment.In order to ensure the correctness of the business logic of each functional module in the system and the stability and ease of use of the system,the unit test code is written after the functional development is completed,and the functional test can be performed according to the business requirements in the project deployment,and finally JMeter is used to test the system.pressure test.