ADAPTIVE DIGITAL ASSISTANT AND SPOKEN GENOME

• US 20180315414A1
• Filed: 04/26/2017
• Published: 11/01/2018
• Est. Priority Date: 04/26/2017
• Status: Abandoned Application

• Current Assignee: Beijing Zitiao Network Technology Company Limited (ByteDance Ltd.)
• Original Assignee: International Business Machines Corporation
• Inventors: Hardee, Christopher J., Joroff, Steve, Nesbitt, Pamela A., Schneider, Scott E.
• Application Number:
  US15/497,536
• Publication Number:
  US 20180315414A1
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/41 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/433 using audio data
• CPC Class Codes: G06F 16/435 Filtering based on additional data, e.g. user or group profiles
• 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 13/0335 Pitch control
• CPC Class Codes: G10L 13/047 Architecture of speech synthesisers
• CPC Class Codes: G10L 13/08 Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command
• CPC Class Codes: G10L 25/63 for estimating an emotional state

Abstract

Embodiments of the invention include a context sensitive adaptive digital assistant for personalized interaction. Embodiments of the invention also include a spoken genome for characterization and analysis of human voice. Aspects of the invention include selecting a starter vocabulary, receiving voice communications from a user, and modifying the starter vocabulary to generate a personalized lexicon. Aspects of the invention also include analyzing and categorizing human voice according to a plurality of characteristics, and creating a spoken genome database.

Claims

1. A computer-implemented method for personalized digital interaction comprising:

selecting a starter vocabulary from a starter vocabulary set;

receiving a plurality of user voice communications from a user;

generating a frequent word list based at least in part on the plurality of user voice communications;

modifying the starter vocabulary with a plurality of words from the frequent word list to generate a personalized lexicon; and

generating a personalized verbal output based at least in part on the personalized lexicon.

2. The method of claim 1, further comprising receiving a secondary learning input from a smart device and modifying the personalized vocabulary based at least in part on the secondary learning input.

3. The method of claim 1, wherein selecting the starter vocabulary comprises:

receiving a user demographic data set,

comparing the user demographic data set to a plurality of characteristics of the starter vocabulary set, and

determining, based at least in part on the comparison, a preferred starter vocabulary.

4. The method of claim 1, further comprising receiving a voice command from the user.

5. The method of claim 4, further comprising:

analyzing the voice command to determine a mood; and

adjusting a characteristic of the personalized verbal output based at least in part on the mood.

6. The method of claim 1, further comprising selecting a specialized vocabulary and augmenting the personalized lexicon with the specialized vocabulary.

7. The method of claim 1, further comprising selecting a set of accent features and adjusting a characteristic of the personalized verbal output based at least in part on the accent features.

8. The method of claim 1, wherein modifying the starter vocabulary comprises replacing a word from the starter vocabulary with a frequent word from the frequent word list or augmenting the starter vocabulary with a new word from the frequent word list.

9. The method of claim 1, wherein the personalized verbal output comprises a personalized gender, pitch, cadence, tonality, rhythm, accent, timing, slurring, elocution, or combination thereof.

10. The method of claim 1, further comprising communicating with an external classification system.

11. A computer program product for characterization and analysis of human voice, wherein the computer program product comprises:

a computer readable storage medium readable by a processing circuit and storing program instructions for execution by the processing circuit for performing a method comprising;

receiving a plurality of media files, wherein the plurality of media files comprise spoken words;

categorizing the plurality of media files according to spoken genome properties to create a categorized spoken genome database;

receiving a user media preference;

determining a user media profile based at least in part on the user media preference, wherein the media profile comprises a spoken genome property; and

providing a media recommendation based at least in part on the user media profile and the categorized spoken genome database.

12. The computer program product of claim 11, wherein the spoken genome property is selected from the group consisting of gender, pitch, cadence, tonality, rhythm, accent, timing, slurring, and elocution.

13. The computer program product of claim 11, wherein the media recommendation comprises a recommended television program, a recommended movie, or a recommended audio book.

14. A processing system for characterization and analysis of human voice comprising:

a processor in communication with one or more types of memory, the processor configured to;

receive a first reference voice sampling, wherein the first reference voice sampling comprises a plurality of reference voices corresponding to a first known reference quality; and

analyze the first voice sampling to determine a spoken genome property corresponding to the first known reference quality.

15. The processing system of claim 14, wherein the processor is configured to output the spoken genome property corresponding to the first known reference quality.

16. The processing system of claim 14, wherein the processor is configured to:

receive a second reference voice sampling, wherein the second reference voice sampling comprises a plurality of reference voices corresponding to a second known reference quality; and

analyze the second voice sampling to determine a spoken genome property corresponding to the second known reference quality.

17. The processing system of claim 16, wherein the processor is further configured to generate a spoken genome database, wherein the spoken genome database comprises:

a first bin comprising the first known reference quality and the spoken genome property corresponding to the first known reference quality, and

a second bin comprising the second known reference quality and the spoken genome property corresponding to the second known reference quality.

18. The processing system of claim 14, wherein the processor is further configured to:

receive a target voice input corresponding to a candidate;

determine a spoken genome property corresponding to the target voice input;

compare the target voice input spoken genome property to the spoken genome property corresponding to the first known reference quality; and

determine whether the candidate corresponds to the first known reference quality.

19. The processing system of claim 18, wherein the processor is configured to output the determination of whether the candidate corresponds the first known reference quality to a display.

20. The processing system of claim 14, wherein the spoken genome property is selected from the group consisting of pitch, cadence, tonality, rhythm, and timing.