Research And Implementation For Intelligent Conversation Robot Based On Microsoft Cognitive Service

Intelligent dialogue robots are one of the most popular technologies nowadays.At home and abroad,many enterprises have introduced dialogue robots to serve their customers,which are used to improve user experience and reduce the cost of employing people.For example,Taobao's smart speaker "sky cat Elf' can not only listen to songs and news,but also use voice dialogue to place orders to buy things.In recent years,more and more enterprises in the face of the future development trend of e-commerce,also began to discuss the introduction of this technology to their services.Businessmen often hesitate to design their own dialogue robots because they are not familiar with the technology.I hope that the experience of this article can provide enterprises preparing to import dialogue robots in the future as a pre-evaluation.This article introduces the background of the company's hope to introduce an intelligent dialogue robot to replace the traditional customer service center for dealing with the company's internal staff consultation some of the more repetitive problems.Dialogue robots use Microsoft Cognitive Services Language Understanding Services and Robot Framework to quickly complete the construction of semantic corpus and question-and-answer interface,which also involves the construction of robot response mode and content database.Finally,in the process of the operation of the dialogue robot,when the accuracy of recognition is low due to the excessive content and the diversification of users' question methods,some improvement measures are proposed,such as hierarchical semantic understanding structure and fuzzy query according to intention,which further improves the correctness of the response.Through the research of this thesis,we can solve the following problems:First,how to classify the data and quickly build a semantic understanding knowledge base;Second,how to build a dialogue robot framework;Third,how to connect the database data in series and follow-up notes and benefit evaluation in the operation process.