System and method for data-driven socially customized models for language generation

• US 10,665,226 B2
• Filed: 06/04/2019
• Issued: 05/26/2020
• Est. Priority Date: 05/13/2014
• Status: Active Grant

• Current Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: Mishra, Taniya, Conkie, Alistair D., Stoyanchev, Svetlana
• Primary Examiner(s): Patel, Shreyans A
• Application Number:
  US16/430,881
• Publication Number:
  US 20190287516A1
• Time in Patent Office: 357 Days
• CPC Class Codes: G10L 13/027 Concept to speech synthesisers; Generation of natural phrases from machine-based concepts generation of parameters for speech synthesis out of text G10L13/08
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/07 to the speaker
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 17/00 Speaker identification or verification techniques
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces

Abstract

Systems, methods, and computer-readable storage devices for generating speech using a presentation style specific to a user, and in particular the user'"'"'s social group. Systems configured according to this disclosure can then use the resulting, personalized, text and/or speech in a spoken dialogue or presentation system to communicate with the user. For example, a system practicing the disclosed method can receive speech from a user, identify the user, and respond to the received speech by applying a personalized natural language generation model. The personalized natural language generation model provides communications which can be specific to the identified user.

Claims

1. A method comprising:

receiving social media postings;

generating, via a processor, a personalized natural language generation model based on a stylistic analysis of the social media postings, wherein the stylistic analysis identifies connections between two or more of a personality independent quotation lattice, a personality independent attribute, a personality dependent attribute, and a user associated with the social media postings; and

applying the personalized natural language generation model to a spoken dialog system.

2. The method of claim 1, further comprising:

identifying, via a processor configured to perform speech analysis, an identity of a user based on characteristics of received speech during a dialog between the user and the spoken dialog system, to yield a user identification.

3. The method of claim 2, wherein the generating of the personalized natural language generation model is further based on the user identification.

4. The method of claim 1, further comprising:

applying the personalized natural language generation model to the spoken dialog system while performing, as part of a dialog, one of automatic speech recognition or natural language generation.

5. The method of claim 1, wherein stylistic analysis further comprises:

identifying speakers in the social media postings, to yield identified speakers;

attributing quoted utterances in the social media postings to the identified speakers, to yield a quotation lattice;

identifying the personality independent attribute and the personality dependent attribute of the quoted utterances within the quotation lattice; and

organizing the quotation lattice based on the personality independent attribute, to yield the personality independent quotation lattice.

6. The method of claim 1, wherein the stylistic analysis further comprises identifying a lexical difference between quotations associated with distinct demographics of speakers within the social media postings.

7. The method of claim 1, wherein the stylistic analysis further comprises identifying a semantic difference between quotations associated with distinct demographics of speakers within the social media postings.

8. The method of claim 1, wherein the stylistic analysis comprises identifying both a lexical difference and a semantic difference between quotations associated with distinct demographics of speakers within the social media postings.

9. The method of claim 2, wherein the identifying of the user from the received speech further comprises performing a prosody analysis of the received speech.

13. The system of claim 1, wherein the computer-readable storage medium stores further instructions stored which, when executed by the processor, cause the processor to perform operations further comprising:

applying the personalized natural language generation model to the spoken dialog system while performing, as part of a dialog, one of automatic speech recognition or natural language generation.

14. The system of claim 1, wherein stylistic analysis further comprises:

identifying speakers in the social media postings, to yield identified speakers;

attributing quoted utterances in the social media postings to the identified speakers, to yield a quotation lattice;

identifying the personality independent attribute and the personality dependent attribute of the quoted utterances within the quotation lattice; and

organizing the quotation lattice based on the personality independent attribute, to yield the personality independent quotation lattice.

10. A system comprising:

a processor; and

a computer-readable storage medium having instructions stored which, when executed by the processor, cause the processor to perform operations comprising;

receiving social media postings;

generating a personalized natural language generation model based on a stylistic analysis of the social media postings, wherein the stylistic analysis identifies connections between two or more of a personality independent quotation lattice, a personality independent attribute, a personality dependent attribute, and a user associated with the social media postings; and

applying the personalized natural language generation model to a spoken dialog system.

11. The system of claim 10, the computer-readable storage medium further stores instructions stored which, when executed by the processor, cause the processor to perform operations further comprising:

identifying an identity of a user based on characteristics of received speech during a dialog between the user and the spoken dialog system, to yield a user identification.

12. The system of claim 11, wherein the generating of the personalized natural language generation model is further based on the user identification.

15. The system of claim 10, wherein the stylistic analysis further comprises identifying a lexical difference between quotations associated with distinct demographics of speakers within the social media postings.

16. The system of claim 10, wherein the stylistic analysis further comprises identifying a semantic difference between quotations associated with distinct demographics of speakers within the social media postings.

17. The system of claim 10, wherein the stylistic analysis comprises identifying both a lexical difference and a semantic difference between quotations associated with distinct demographics of speakers within the social media postings.

18. The system of claim 11, wherein the identifying of the user from the received speech further comprises performing a prosody analysis of the received speech.

19. A non-transitory computer-readable storage device having instructions stored which, when executed by a computing device, cause the computing device to perform operations comprising:

receiving social media postings;

generating a personalized natural language generation model based on a stylistic analysis of the social media postings, wherein the stylistic analysis identifies connections between two or more of a personality independent quotation lattice, a personality independent attribute, a personality dependent attribute, and a user associated with the social media postings; and

applying the personalized natural language generation model to a spoken dialog system.

20. The non-transitory computer-readable storage device of claim 19, wherein the non-transitory computer-readable storage device stores further instructions stored which, when executed by the computing device, cause the computing device to perform operations further comprising:

identifying an identity of a user based on characteristics of received speech during a dialog between the user and the spoken dialog system, to yield a user identification.