Bus multi-mode question and answer assistant construction method, system, platform and customer service robot

Abstract

The embodiment of the invention discloses a method, a system, a platform and a customer service robot for constructing a bus multi-modal question and answer assistant. The method comprises the following steps: collecting an instruction input by a user to obtain a multi-modal data sample set; creating a multi-modal fusion model and obtaining a structured query intention corresponding to an instruction input by a user in real time; creating a plurality of types of multi-modal question and answer digital person assistants, determining whether the multi-modal question and answer digital person assistants need to be called according to the structured query intention, if so, performing content reply through the determined type of multi-modal question and answer digital person assistants, and if not, calling a preset chat digital person assistant to perform content reply; the invention designs a public transportation industry intelligent question-answering technology fusing multi-modal perception, a dynamic knowledge base and context reasoning, and can effectively solve the problems of multi-modal information splitting, poor dynamic adaptability, semantic ambiguity, insufficient anti-interference capability and the like of a traditional question-answering system in a public transportation scene.

Images