Research On The Application Of Intelligent Dialogue System In Improving The Operational Efficiency Of Government Service

In recent years,the knowledge economy and virtual society have developed rapidly,and the public government service demand gradually presents the characteristics of informatization and convenience.However,on the service side,the traditional method of relying on manual reply has obvious limitations,such as consultation waiting,weak ability of cross department information integration,human fatigue,and unstable personnel.In order to meet the public needs,improve service efficiency and improve service quality,governments have generally strengthened the construction of e-government service based on Internet and artificial intelligence technology.With the rapid development of deep learning,reinforcement learning and other basic algorithms,the overall architecture,algorithm system and application mode of human-computer dialogue system have been greatly changed and improved.Traditionally,dialogue system is an information retrieval method based on natural language,which accepts questions in natural language and outputs the only accurate or appropriate answers.In recent years,the interaction mode of dialogue system has been further expanded from natural language to voice,image,gesture,expression and other modes,and the output has also been expanded from answers in natural language to multimedia messages Information and services.Dialog system has been widely used in many scenarios in the field of e-government,such as intelligent customer service,intelligent outbound call,intelligent voice quality inspection,intelligent training partner,intelligent assistant,etc.These applications have not only improved service efficiency and quality,but also exposed several problems.Firstly,for specific fields such as government services,there are characteristics of strong professional knowledge and frequent business updates,so it is difficult to obtain large-scale effective data to build an end-to-end dialogue system based on neural network.Second,most of the existing dialogue system based on knowledge base entry depends on the professional knowledge,low input efficiency,the context semantic problems such as difficult to obtain,and the existing knowledge in many-to-many entity relation extraction method,event extraction in areas such as there is still a sore point,cause existing dialogue system based on the existing knowledge of the problem is difficult to achieve effective semantic understandingThis paper first studied the government affairs service center operating efficiency indicators,and further on the dialogue system in the government affairs service a variety of scenarios are studied,analyzed the problems and challenges.After that,this paper proposes a dialogue system architecture design based on knowledge extraction,and analyzes the specific characteristics of the architecture applied to government services.