Confidence level assignment to information from audio transcriptions

• US 9,002,702 B2
• Filed: 05/03/2012
• Issued: 04/07/2015
• Est. Priority Date: 05/03/2012
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Bostick, James E., Ganci, John M. Jr., Kaemmerer, John P., Trim, Craig M.
• Primary Examiner(s): SPOONER, LAMONT M
• Application Number:
  US13/463,106
• Publication Number:
  US 20130297291A1
• Time in Patent Office: 1,069 Days
• Field of Search: 704/1, 704/9, 704/10, 704/235, 704/246, 707706-708
• US Class Current: 704/9
• CPC Class Codes: G06F 40/157 using dictionaries or tables
• CPC Class Codes: G06F 40/189 Automatic justification
• CPC Class Codes: G06F 40/226 Validation
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06F 40/268 Morphological analysis
• CPC Class Codes: G06F 40/51 Translation evaluation
• CPC Class Codes: G10L 15/01 Assessment or evaluation of speech recognition systems
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 17/20 Pattern transformations or operations aimed at increasing system robustness, e.g. against channel noise or different working conditions

Abstract

Embodiments of the present invention provide an approach for automatically assigning a confidence level to information extracted from a transcription of a voice recording. Specifically, in a typical embodiment, an axiom is extracted from a source associated with the text of the transcription. A confidence level of the source is determined. A confidence level is assigned to the axiom based on the confidence level of the source.

Claims

1. A method of assigning a confidence level to an axiom extracted from a text of a transcription of a voice recording;

receiving, at a computer device, the text of the transcription;

comparing, using the computer device, every word from the text to a customer specific dictionary and a dictionary of common language words;

determining a number of inaccurately spelled words in the transcription based on the comparing;

assigning, using a Gaussian distribution, a confidence level to the text of the transcription based on the determining;

estimating an accuracy of the text based on the assigned confidence level;

gathering, using the computer device, external information related to words in the text by retrieving a set of axioms that further define the words in the text of the transcription from a set of sources, each axiom in the set of axioms comprising a proposition that is regarded as being at least one of established, accepted, and self-evidently true;

determining a confidence level of each source of the set of sources; and

assigning a confidence level to each axiom of the set of axioms based on a combination of the confidence level of the set of sources and the accuracy of the text estimated based on the assigned confidence level.

2. The method of claim 1, further comprising associating a confidence level to a plurality of sources prior to receiving the text of the transcription.

3. The method of claim 1, wherein the axiom is stored in a knowledge base.

4. The method of claim 1, further comprising generating a notification indicating the assigned confidence level of the axiom.

5. The method of claim 1, further comprising interpreting the text of the transcription using natural language processing (NLP).

6. A system for assigning a confidence level to an axiom extracted from a text of a transcription of a voice recording, comprising:

a memory medium comprising instructions;

a bus coupled to the memory medium; and

an audio transcription tool coupled to the bus that when executing the instructions causes the system to;

receive the text of the transcription;

compare every word from the text to a customer specific dictionary and a dictionary of common language words;

determine a number of inaccurately spelled words in the transcription based on the comparing;

assign, using a Gaussian distribution, a confidence level to the text of the transcription based on the determining;

estimate an accuracy of the text based on the assigned confidence level;

gather external information related to words in the text by retrieving a set of axioms that further define the words in the text of the transcription from a set of sources, each axiom in the set of axioms comprising a proposition that is regarded as being at least one of established, accepted, and self-evidently true;

determine a confidence level of each source of the set of sources; and

assign a confidence level to each axiom of the set of axioms based on a combination of the confidence level of the set of sources and the accuracy of the text estimated based on the assigned confidence level.

7. The system of claim 6, wherein the axiom is stored in a knowledge base.

8. The system of claim 6, the memory medium further comprising instructions for causing the system to generate a notification indicating the assigned confidence level of the axiom.

9. The system of claim 6, wherein the axiom is stored in a knowledge base.

10. The system of claim 6, the memory medium further comprising instructions for causing the system to interpret the text of the transcription using natural language processing

11. The system of claim 6, the memory medium further comprising instructions for causing the system to estimate an accuracy of the text of the transcription based on the confidence level of the source.

12. A computer program product for assigning a confidence level to an axiom extracted from a text of a transcription of a voice recording, the computer program product comprising a non-transitory computer readable storage medium, and program instructions stored on the non-transitory computer readable storage medium, to:

receive the text of the transcription;

compare every word from the text to a customer specific dictionary and a dictionary of common language words;

determine a number of inaccurately spelled words in the transcription based on the comparing;

assign, using a Gaussian distribution, a confidence level to the text of the transcription based on the determining;

estimate an accuracy of the text based on the assigned confidence level;

gather external information related to words in the text by retrieving a set of axioms that further define the words in the text of the transcription from a set of sources, each axiom in the set of axioms comprising a proposition that is regarded as being at least one of established, accepted, and self-evidently true;

determine a confidence level of each source of the set of sources; and

assign a confidence level to each axiom of the set of axioms based on a combination of the confidence level of the set of sources and the accuracy of the text estimated based on the assigned confidence level.

13. The computer program product of claim 12, the non-transitory computer readable storage medium further comprising instructions to associate a confidence level to a plurality of sources prior to receiving the text of the transcription.

14. The computer program product of claim 12, wherein the axiom is stored in a knowledge base.

15. The computer program product of claim 12, the non-transitory computer readable storage medium further comprising instructions for causing the system to generate a notification indicating the assigned confidence level of the axiom.

16. The computer program product of claim 12, the non-transitory computer readable storage medium further comprising instructions to interpret the text of the transcription using natural language processing (NLP).

17. A method for deploying a system for assigning a confidence level to an axiom extracted from a text of a transcription of a voice recording, comprising:

providing a computer infrastructure being operable to;

receive, at a computer device, the text of the transcription;

compare, using the computer device, every word from the text to a customer specific dictionary and a dictionary of common language words;

determine a number of inaccurately spelled words in the transcription based on the comparing;

assign, using a Gaussian distribution, a confidence level to the text of the transcription based on the determining;

estimate an accuracy of the text based on the assigned confidence level;

gather, using the computer device, external information related to words in the text by retrieving a set of axioms that further define the words in the text of the transcription from a set of sources, each axiom in the set of axioms comprising a proposition that is regarded as being at least one of established, accepted, and self-evidently true;

determine a confidence level of each source of the set of sources; and

assign a confidence level to each axiom of the set of axioms based on a combination of the confidence level of the set of sources and the accuracy of the text estimated based on the assigned confidence level.

18. The method for deploying a system of claim 17, further comprising providing a computer infrastructure being operable to estimate an accuracy of the text of the transcription based on the confidence level of the source.