Speech communication system and method with human-machine coordination

A speech communication system with human-machine coordination, comprising a communications module, and further comprising:

• an answering and calling-out module, allocating a human agent after a call is connected;
• a speech intention understanding module, comprising a speech intention understanding model, for understanding an intention in a call process;
• a voice cloning module, comprising a voice cloning model, for synthesizing voice of a corresponding attendant from a reply text;
• a human-machine interaction module, invoking the speech intention understanding module through a speech robot to obtain an intention of a client and further invoking, according to a pre-edited speechcraft, the voice cloning module to communicate with the client by the voice of the corresponding attendant;
• an intervention prompting module, comprising a quantization analysis submodule, monitoring in real-time call content during a human-machine interaction to obtain an intervention willing probability and prompting the human agent according to a predefined rule, wherein the intervention willing probability is a probability in which a prospective client is generated;
• if the intervention willing probability is greater than a preset threshold, the human agent will be invoked to intervene in the call;
• and
• a manual intervention module, replying the client by speech of the corresponding attendant and/or by processing reply content of a random attendant into voice of the corresponding attendant through invoking the voice cloning module, and further comprising a retrieving and prompting submodule that can invoke the speech intention understanding module, to display a call intention and give the random attendant a prompt for conversation.