Bond marker automatic detection technology based on deep learning

The invention provides a bond mark automatic detection technology based on deep learning, and relates to the technical field of financial AI. The bond mark automatic detection technology based on the deep learning comprises the following specific modules: a universal interface module for receiving information provided by a chat system and performing simple processing; a text analysis and recognition module for recognizing intentions of 'bidding, bidding changing, bidding withdrawing, newly adding' and the like of the user and extracting corresponding elements; a reforming module for organizing into a standard data form based on information provided by a user and records in a historical database; an execution module for executing specific operations such as 'bidding, bid changing and bid withdrawing'; and a memory module for storing bidding information of the user. The technology has the beneficial effects that: the user intentions of 'bidding/revising/withdrawing/adding' and the like in the chat are automatic