Application of task-oriented dialogue in operations for future networks to business

At present, the network operation services provided by operators for to business (toB) customers have customer friendliness issues mostly affected by manual input. The limitations are reflected in four aspects: time consumption and understanding differences in the process of customer intention communication, waiting time and lack of process transparency. Therefore, in order to provide a more automatic, standard, concise and user-engaged method on operations for future networks in the toB field, this paper proposes applying a task-oriented dialogue system to the network operation task processing. The most significant advantage is that it makes possible for industry customers to participate directly in and facilitate their task completion process together with the system by several turns of dialogue. The conventional pipeline structure which combines both neural networks and rules is appropriate for meeting the above requirements for providing unified natural language understanding of intent, tracking dialogue state and generating actions of reply or command. With the help of a knowledge database and a network management system, the dialogue system provides users with more detailed operation information, which directly promotes user participation and transparency of the whole service process.