Intelligent conversational systems
A system and method simulate conversation with a human user. The system and method receive media, convert the media into a system-specific format, and compare the converted media to a vocabulary. The system and method generate a plurality of intents and a plurality of sub-entities and transform them into a pre-defined format. The system and method route intents and the sub-entities to a first selected knowledge engine and a second knowledge engine. The first selected knowledge engine selects the second knowledge engine and each active grammar in the vocabulary uniquely identifies each of the knowledge engines.
1. A system that simulates conversation with a human user, comprising:
a speech engine that processes a plurality of spoken utterances received from the human user into a plurality of representations;
the speech engine compares the plurality of representations to a first vocabulary to render a speech output comprising a plurality of interpretations of the plurality of spoken utterances, process commands, and intents;
a recognition processor that receives the speech output and converts the speech output into a system-specific format;
a first natural language processing engine that compares the speech output to a second vocabulary and generates a plurality of second intents and a plurality of sub-entities; and
a controller that transforms the plurality of second intents and the plurality of sub-entities into a pre-defined format and routes the plurality of second intents and the plurality of sub-entities to a first selected knowledge engine and a second knowledge engine;
where the first selected knowledge engine selects the second knowledge engine based on a knowledge base and a downloadable profile;
where the first selected knowledge engine or the second knowledge engine engages with the human user through one or more dialogues based on a personality type of a fictional character; and
where the one or more dialogues is rendered through an acoustic modeler and a voice generator that deliver a natural-sounding voice through a plurality of audio samples.
2. The system of claim 1 where the one or more dialogues are indistinguishable from human speech.
3. The system of claim 2 where the system converses in an auditory tone that comprise human speech.
4. The system of claim 3 where the personality type is rendered through a customized vocabulary file that models traits of the fictional character.
5. The system of claim 1 where the personality type is rendered through a customized vocabulary file that models traits of the fictional character.
6. The system of claim 1 where the personality type is rendered through a customized vocabulary file and a programmable logic selectable by the recognition processor that models traits of the fictional character.
7. The system of claim 1 where the dialogue is rendered from a customized vocabulary and a programming logic selectable by the first and the second knowledge engine.
8. The system of claim 1 where the dialogue is rendered from a customized vocabulary and a programming logic selectable by the controller.
9. The system of claim 1 further comprising an attitude classifier that modifies the one or more dialogues in response to an attitude classifier that rates a mental disposition of a user by analyzing the plurality of spoken utterances.
10. The system of claim 9 where the attitude classifier classifies in configured to classify the plurality of spoken utterances as expressing a positive emotion, a neutral emotion, a negative emotion, and a mixed emotion.
11. The system of claim 1 further comprising an exchange manager that updates the first vocabulary through a regulated process that tags and classifies an input when the one or more dialogues comprising an input has a rating value greater than a smoothed predetermined value.
12. The system of claim 1 where the first selected knowledge engine and the second knowledge engine comprise an insurance quoting knowledge engine, a claims processing knowledge engine, or an on-line insurance servicing knowledge engine.
13. The system of claim 1 further comprising a second natural language processing engine that compares an output from the recognition processor to a third vocabulary and generates a third plurality of intents and a second plurality of sub-entities.
14. A method that simulates conversation with a human user, comprising:
processing a plurality of spoken utterances received from the human user to a plurality of representations through a speech engine;
comparing the plurality of representations to a first vocabulary to render an output comprising a plurality of interpretations of the plurality of spoken utterances, process commands, and intents;
receiving the output and converting the output into a system-specific format at a recognition processor;
comparing the output to a second vocabulary and generating a plurality of second intents and a plurality of sub-entities at a first natural language processing engine; and
transforming the plurality of second intents and the plurality of sub-entities into a pre-defined format, and routing the plurality of second intents and the plurality of sub-entities to a first selected knowledge engine and a second knowledge engine at a controller;
where the first selected knowledge engine selects the second knowledge engine based on a knowledge base and a downloadable profile;
where the first selected knowledge engine or the second knowledge engine engages with the human user through one or more dialogues based on a personality type of a fictional character; and
where the one or more dialogues is rendered through an acoustic modeler and a voice generator that deliver a natural-sounding voice through a plurality of audio samples.
15. The method of claim 14 where the one or more dialogues are indistinguishable from human speech.
16. The method of claim 15 where the method tracks the one or more dialogues through an exchange manager and updates the first vocabulary through a backpropagation based on the one or more dialogues and a virtual assistant.
17. The method of claim 16 where the exchange manager and the virtual assistant monitor and manage a plurality of physically separated and logically separated networks.
18. The method of claim 17 where virtual assistant renders a landing page providing information on each of the plurality of physically separated and logically separated networks.
19. The method of claim 18 where the virtual assistant includes a filter that sorts the information based on a relevance criteria that measure key words and an occurrence of a plurality of weighted key words.
20. The method of claim 18 where the virtual assistant comprises a fully automated regulated system that is rule based.
21. The method of claim 16 where the virtual assistant comprises a neural network.
22. The method of claim 14 further comprising a knowledge base accessible to the first selected knowledge engine or the second knowledge engine that includes a plurality of attributes that identify a human user's intention.
23. The method of claim 14 further comprising an attitude classifier that modifies the one or more dialogues in response to an attitude classifier that rates a mental disposition of a user by analyzing the plurality of spoken utterances.
24. The method of claim 14 where the second vocabulary comprises active grammars that include a plurality of words and phrases to be recognized.
25. The method of claim 14 where the first selected knowledge engine and the second knowledge engine comprise an insurance quoting knowledge engine, a claims processing knowledge engine, or an on-line insurance servicing knowledge engine.
26. The method of claim 14 further comprising a second natural language processing engine that compares an output from the recognition processor to a third vocabulary and generates a third plurality of intents and a second plurality of sub-entities.
27. A non-transitory machine-readable medium encoded with machine-executable instructions, where execution of the machine-executable instructions is for:
processing a plurality of spoken utterances received from a human user to a plurality of representations through a speech engine;
comparing the plurality of representations to a first vocabulary to render an output comprising a plurality of interpretations of the plurality of spoken utterances, process commands, and intents;
receiving the output and converting the output into a system-specific format at a recognition processor;
comparing the output to a second vocabulary and generating a plurality of second intents and a plurality of sub-entities at a first natural language processing engine; and
transforming the plurality of second intents and the plurality of sub-entities into a pre-defined format and routing the plurality of second intents and the plurality of sub-entities to a first selected knowledge engine and a second knowledge engine at a controller;
where the first selected knowledge engine selects the second knowledge engine based on a knowledge base and a downloadable profile; and
where the first knowledge engine or the second knowledge engine engages with the human user through one or more dialogues based on a personality type of a fictional character and is updated by a virtual manager through a backpropagation;
where the one or more dialogues is rendered through an acoustic modeler and a voice generator that deliver a natural-sounding voice through a plurality of audio samples.
28. The non-transitory machine-readable medium of claim 27 where the one or more dialogues are indistinguishable from human speech.
29. The non-transitory machine-readable medium of claim 28 where the non-transitory machine-readable medium renders an output that converses in an auditory tone that is indistinguishable from human speech.
30. The non-transitory machine-readable medium of claim 27 further comprising an attitude classifier that modifies the one or more dialogues in response to an attitude classifier that rates a mental disposition of a user by analyzing the plurality of spoken utterances.