Intelligent conversational systems

• US 10,305,826 B1
• Filed: 05/03/2018
• Issued: 05/28/2019
• Est. Priority Date: 05/03/2018
• Status: Active Grant

• Current Assignee: Progressive Casualty Insurance Company (Progressive Corporation), Progressive Insurance Company
• Original Assignee: Progressive Casualty Insurance Company (Progressive Corporation)
• Inventors: White, Matthew T., Surtz, Brian J., Cox, Callen C.
• Primary Examiner(s): Smith, Paulinho E
• Application Number:
  US15/970,632
• Time in Patent Office: 390 Days
• CPC Class Codes: G06F 40/56 Natural language generation
• CPC Class Codes: G06N 3/006 based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06N 5/041 Abduction
• CPC Class Codes: G06Q 40/08 Insurance
• CPC Class Codes: G10L 13/00 Speech synthesis; Text to speech systems
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: H04L 51/02 using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages

Abstract

A system and method simulates 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.

Claims

1. A system that simulates conversation with a human user, comprising:

a recognition processor that receives media and converts the media into a system-specific format;

a first natural language processing engine that compares the converted media to a vocabulary and generates a plurality of intents and a plurality of sub-entities; and

a controller that transforms the plurality of intents and the plurality of sub-entities into a pre-defined format and routes the plurality of 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; and

where each active grammar in the vocabulary uniquely identifies each of the plurality of knowledge engines.

2. The system of claim 1 further comprising a knowledge base accessible to the first selected knowledge engine and the second knowledge engine that includes a plurality of attributes that identify a user'"'"'s intention.

3. The system of claim 2 where the knowledge base comprises a plurality of downloadable profiles.

4. The system of claim 1 where the active grammars comprise a plurality of words and phrases to be recognized.

5. 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.

6. The system of claim 1 further comprising a second natural language processing engine that compares a second converted media to the vocabulary and generates a second plurality of intents and a second plurality of sub-entities.

7. The system of claim 6 where the first and the second natural processing engines comprise an automatic speech recognition service.

8. A method that simulates conversation with a human user, comprising:

receiving media and converting the media into a system-specific format;

comparing the converted media to a vocabulary and generating a plurality of intents and a plurality of sub-entities; and

transforming the plurality of intents and the plurality of sub-entities into a pre-defined format and routing the plurality of 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; and

where each active grammar in the vocabulary uniquely identifies each of the plurality of knowledge engines.

9. The method of claim 8 further comprising accessing a plurality of attributes that identify a user'"'"'s intention.

10. The method of claim 9 where the plurality of attributes are stored in a plurality of downloadable profiles.

11. The method of claim 8 where the active grammars comprise a plurality of words and phrases to be recognized.

12. The method of claim 8 where the first selected knowledge engine and the second knowledge engine comprise an insurance quoting knowledge engine, a claims processing knowledge engine, and an on-line insurance servicing knowledge engine.

13. The method of claim 8 where an exchange manager identifies a plurality of words or a phrase associated with the identified user'"'"'s generation and renders the plurality of words or the phrase on a social media device in response to the received media.

14. The method of claim 8 further comprising comparing a second converted media to the vocabulary and generating a second plurality of intents and a second plurality of sub-entities.

15. The method of claim 14 where the first and the second plurality of intents and the first and the second plurality of sub-entities identify speech.

16. A non-transitory machine-readable medium encoded with machine-executable instructions, where execution of the machine-executable instructions is for:

receiving media and converting the media into a system-specific format;

comparing the converted media to a vocabulary and generating a plurality of intents and a plurality of sub-entities; and

transforming the plurality of intents and sub-entities into a pre-defined format and routing the plurality of 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; and

where each active grammar of the vocabulary uniquely identifies each of the plurality of knowledge engines.

17. The non-transitory machine-readable medium of claim 16 further comprising accessing a plurality of attributes that identify a user'"'"'s intention.

18. The non-transitory machine-readable medium of claim 17 where the plurality of attributes are stored in a plurality of downloadable profiles.

19. The non-transitory machine-readable medium of claim 16 where the active grammars comprise a plurality of words and phrases to be recognized.

20. The non-transitory machine-readable medium of claim 16 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.

21. The non-transitory machine-readable medium of claim 16 further comprising comparing a second converted media to the vocabulary and generating a second plurality of intents and a second plurality of sub-entities.

22. The non-transitory machine-readable medium of claim 21 where the first and second plurality of intents and the first and second plurality of sub-entities identify speech.

23. A non-transitory machine-readable medium encoded with machine-executable instructions, where execution of the machine-executable instructions is for:

receiving media and converting the media into a system-specific format;

comparing the converted media to a vocabulary and generating a plurality of intents and a plurality of sub-entities; and

transforming the plurality of intents and the plurality of sub-entities into a pre-defined format and routing the plurality of intents and the plurality of sub-entities via a controller to a first selected knowledge engine and a second knowledge engine; and

identifying a user'"'"'s generation through an exchange engine in communication with the controller;

where the first selected knowledge engine selects the second knowledge engine and each active grammar of the vocabulary uniquely identifies each of the plurality of knowledge engines.

24. The non-transitory machine-readable medium of claim 23 further comprising accessing a plurality of attributes that identify a user'"'"'s intention.

25. The non-transitory machine-readable medium of claim 24 where the plurality of attributes are stored in a plurality of downloadable profiles.

26. The non-transitory machine-readable medium of claim 23 where the active grammars comprise a plurality of words or phrases to be recognized.

27. The non-transitory machine-readable medium of claim 23 where the first selected knowledge engine and the second knowledge engines comprise at least two separate engines selected from a group comprising:

an insurance quoting knowledge engine, a claims processing knowledge engine, or an on-line insurance servicing knowledge engine.

28. The non-transitory machine-readable medium of claim 23 where the exchange manager identifies a plurality of words or a phrase associated with the identified user'"'"'s generation and renderers the plurality of words or phrases on a social media device in response to the received media.

29. The non-transitory machine-readable medium of claim 23 further comprising comparing a second converted media to the vocabulary and generating a second plurality of intents and a second plurality of sub-entities.

30. The non-transitory machine-readable medium of claim 29 where the first and the second plurality of intents and the first and the second plurality of sub-entities identify speech.

31. A non-transitory machine-readable medium encoded with machine-executable instructions, where execution of the machine-executable instructions is for:

receiving media and converting the media into a system-specific format;

comparing the converted media to a vocabulary and generating a plurality of intents and a plurality of sub-entities; and

transforming the plurality of intents and the plurality of sub-entities into a pre-defined format and routing the plurality of intents and the plurality of sub-entities via a controller to a first selected knowledge engine and a second knowledge engine; and

transmitting a location and an identification of a physical object via a vision engine to the controller;

where the first selected knowledge engine selects the second knowledge engine and each active grammar of the vocabulary uniquely identifies each of the plurality of knowledge engines.

32. A system that simulates conversation with a human user, comprising:

a recognition processor that receives media and converts the media into a system-specific format;

a first natural language processing engine that compares the converted media to a vocabulary and generates a plurality of intents and a plurality of sub-entities; and

a controller that transforms the plurality of intents and the plurality of sub-entities into a pre-defined format and routes the plurality of intents and the plurality of sub-entities to a first selected knowledge engine and a second knowledge engine; and

means that identifies the human user'"'"'s generation;

where the first selected knowledge engine selects the second knowledge engine and each active grammar in the vocabulary uniquely identifies each of the plurality of knowledge engines.

33. A system that simulates conversation with a human user, comprising:

a recognition processor that receives media and converts the media into a system-specific format;

a first natural language processing engine that compares the converted media to a vocabulary and generates a plurality of intents and a plurality of sub-entities; and

a controller that transforms the plurality of intents and the plurality of sub-entities into a pre-defined format and routes the plurality of intents and the plurality of sub-entities to a first selected knowledge engine and a second knowledge engine; and

means that transmits a location and an identification of a physical object to the controller;

where the first selected knowledge engine selects the second knowledge engine and each active grammar in the vocabulary uniquely identifies each of the plurality of knowledge engines.