Design And Implementation Of 114 Intelligent Number Query System Based On Multi-Round Dialogue

The task of 114 number search tasks is clear and the process repetition rate is high,but manual customer service needs certain memory and search speed to take into account the accuracy and efficiency of the number search.In the context of today's intelligent social services,the related technologies of artificial intelligence are rapidly developing.Artificial intelligence replaces manual customer service to complete interactive tasks.More and more application scenarios,such as ticket booking,hotel booking,and number search,are now beginning to access Intelligent call system to complete.However,for companies applying artificial intelligence,in addition to hoping to liberate manual customer service through intelligent calls and reduce enterprise costs,how to ensure the sensitivity of the system to new data while continuously updating data during the long-term operation of the intelligent system,and to ensure the availability and stability of the system by monitoring the call situation,is also a problem that needs to be solved for the enterprise.Therefore,the purpose of this topic is to design a usable and intelligent number search system that replaces manual customer service with robots to reduce the workload of manual customer service and at the same time allow companies to manage the number search information more conveniently.The research content of this topic is to design and implement a set of intelligent 114 number search system based on the concept and technology of multi-round dialogue.In addition to replacing the manual voice interaction with the user to complete the number search function,the system has applied to multi-round dialogue management.Based on the technology of slot filling,the user experience and call efficiency are fully considered in the process;for understanding user input,BLSTM(Bi-directional Long Short-Term Memory,bidirectional long-term memory network model)is used for intent recognition,and a bidirectional long-short-term memory model combined with CRF(Conditional Random field,conditional random field model)is used to identify the slot;for enterprise management,a business management system was developed using the express framework based on nodejs,which provides functions such as updating data,managing AI models and files,managing blacklists,and monitoring call information.The thesis expounds the research background and significance of the subject,makes a general survey on the related technical status of the subject,designs the system architecture,selects stable and applicable technologies for each module of the system,and then develops and tests each module.Finally,it is verified that the system meets the needs of its function and performance and has practical use value.