Method, system and apparatus for understanding and generating human conversational cues

A method for understanding and generating human conversational cues, comprising:

• utilizing a conversation intelligence (CI) manager module having a rule-based engine on conversational intelligence for a voice-based digital assistant (VDA) to process information from one or more modules to make determinations on both i) understanding the human conversational cues and ii) generating human-like conversational cues, including understanding and/or generating a backchannel utterance, in a flow and exchange of human communication in order to at least one of 1) take, grab or yield a conversational floor between a user and the VDA, and 2) establish a conversational grounding without taking the conversational floor, utilizing the rule-based engine to analyze and make a determination on a conversational cue of, at least, prosody in a user's speech to generate the backchannel utterance to signal any of i) an understanding, ii) a correction, iii) a confirmation, and iv) a questioning of verbal communications conveyed by the user in the flow of speech, utilizing the rule-based engine to analyze to make a determination on what backchannel utterances to generate based upon linguistically motivated rules scripted out for two or more micro-interaction scenarios, wherein a first micro-interaction scenario is selected from a group consisting of
• I. to use at least one of a tone or a pitch of the VDA to determine whether to take, grab, or yield the conversational floor between the user and the VDA,
• II. to utter a question to establish the conversational grounding but not grab the conversational floor when a sustained pause in a conversation occurs beyond a set time duration, rather than uttering the backchannel utterance to indicate a desire of the VDA to grab the conversational floor,
• III. to adjust a behavior to the VDA to match a behavior of the user when the CI manager module determines the user is distracted,
• IV. to wait for a fixed time period and then issue the backchannel utterance to encourage additional information from the user when the user has not completed their thought,
• V. to issue a backchannel or other quick words to reinforce a point being made when the user generates a user self-correction in the user's speech, VI, to change a pace of how the VDA normally communicates when either the user or the VDA generates either a long list of items or a complex list of items, and
• VII. to generate an acknowledgement and then wait for a further user input when the user generates an indication to hang on to the VDA, and using one or more non-transitory machine-readable media, one or more processors, and where algorithms of the rule-based engine and the CI manager module are implemented in hardware electronic components, software stored in the one or more non-transitory machine-readable media to be executed by the one or more processors, and any combination of both.

An apparatus for understanding and generating human conversational cues, comprising:

• a conversation intelligence (CI) manager module configured to use a rule-based engine configured to use rules and parameters on conversational intelligence for a voice-based digital assistant (VDA), where the CI manager module has one or more inputs to receive information as the parameters from one or more modules to make determinations on both i) understanding the human conversational cues and ii) generating human-like conversational cues, including at least understanding and/or generating a backchannel utterance, in a flow and exchange of human communication in order to at least one of 1) take, grab, or yield a conversational floor between a user and the VDA, and 2) establish a conversational grounding without taking the conversational floor, where the CI manager module is configured to use reinforcement learning using the rules and the parameters to analyze and make a determination on a conversational cue of, at least, prosody in a user's speech, and when the CI manager module decides to generate the backchannel utterance, then the CI manager module is configured to generate a command to issue the backchannel utterance to signal any of i) an understanding, ii) a prompt for further information, iii) a confirmation, and iv) a questioning of verbal communications conveyed by the user in the user's speech, where the rule-based engine is configured to analyze to make a determination on what backchannel utterances to generate based upon linguistically motivated rules scripted out for two or more micro-interaction scenarios, wherein a first micro-interaction scenario is selected from a group consisting of
• I. to use at least one of a tone or a pitch of the VDA to determine whether to take, grab, or yield the conversational floor between the user and the VDA,
• II. to utter a question to establish the conversational grounding but not grab the conversational floor when a sustained pause in a conversation occurs beyond a set time duration, rather than uttering the backchannel utterance to indicate a desire of the VDA to grab the conversational floor,
• III. to adjust a behavior to the VDA to match a behavior of the user when the CI manager module determines the user is distracted,
• IV. to wait for a fixed time period and then issue the backchannel utterance to encourage additional information from the user when the user has not completed their thought,
• V. to issue a backchannel or other quick words to reinforce a point being made when the user generates a user self-correction in the user's speech,
• VI. to change a pace of how the VDA normally communicates when either the user or the VDA generates either a long list of items or a complex list of items, and
• VII. to generate an acknowledgement and then wait for a further user input when the user generates an indication to hang on to the VDA, and one or more non-transitory machine-readable media, one or more processors, and where algorithms of the rule-based engine and the CI manager module are implemented in hardware electronic components, software stored in the one or more non-transitory machine-readable media to be executed by the one or more processors, and any combination of both.