Method and apparatus for activating and deactivating auxiliary topic libraries in a speech dictation system

• US 6,360,201 B1
• Filed: 06/08/1999
• Issued: 03/19/2002
• Est. Priority Date: 06/08/1999
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Lewis, James R., Ortega, Kerry A., Wang, Huifang, Vanbuskirk, Ronald
• Primary Examiner(s): {haeck over (S)}mits, Ta̅livaldis Ivars
• Assistant Examiner(s): Azad, Abul K.
• Application Number:
  US09/328,087
• Time in Patent Office: 1,015 Days
• Field of Search: 704/10, 704/231, 704/236, 704/237, 704/239, 704/243, 704/244, 704/251, 704/252, 704/253, 704/257
• US Class Current: 704/251
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering
• CPC Class Codes: G10L 2015/228 of application context

Abstract

A dictation system (100) performs a method of dictating speech which automatically activates (502) and deactivates (306, 408, 510) auxiliary topic libraries based on the input speech. After receiving (206) input speech, the method searches (208, 214) a general library and topic libraries that are currently active, if any. The method also searches (220, 226) all or portions of inactive topic libraries. If the spoken word is recognized in a particular inactive topic library, the method automatically activates (502) that topic library. In a preferred embodiment, the method maintains an adjustable “score” for each active topic library. An active library'"'"'s score is increased (402) each time a word is recognized in the library, and decreased (302, 404, 506) when a word is recognized in another library. If the score falls below a certain threshold, the active topic library is automatically deactivated (306, 408, 510).

Claims

1. A method for dictating speech using a general library and at least one auxiliary topic library, the method comprising the steps of:

a) receiving a spoken word of input speech;

b) comparing the spoken word to words within one or more active auxiliary topic libraries;

c) if the spoken word is recognized as being one of the words, dictating the spoken word to be the one of the words and returning to the step of receiving;

d) comparing the spoken word to other words within an inactive auxiliary topic library;

e) if the spoken word is recognized as being one of the other words, automatically activating the inactive auxiliary topic library to be an active auxiliary topic library, dictating the spoken word to be the one of the other words, and returning to the step of receiving;

f) determining whether an active auxiliary topic library should be deactivated; and

g) if the active auxiliary topic library should be deactivated, automatically deactivating the active auxiliary topic library.

2. A method for dictating speech using a general library and at least one auxiliary topic library, the method comprising the steps of:

a) receiving a spoken word of input speech;

b) comparing the spoken word to words within one or more active auxiliary topic libraries;

b1) comparing the spoken word to words within a general library;

b2) when word Is In the general library, adjusting scores of other active auxiliary topic libraries;

b3) determining, based on the scores, If any of the other active auxiliary topic libraries should be deactivated;

b4) If any of the other active auxiliary topic libraries should be deactivated, automatically deactivating those of the other active auxiliary topic libraries that should be deactivated;

c) If the spoken word is recognized as being one of the words, dictating the spoken word to be the one of the words and returning to the step of receiving;

d) comparing the spoken word to other words within an inactive auxiliary topic library; and

,

e) if the spoken word is recognized as being one of the other words, automatically activating the inactive auxiliary topic library to be an active auxiliary topic library, dictating the spoken word to be the one of the other words, and returning to the step of receiving.

3. The method as claimed in claim 2, wherein the step of adjusting the scores comprises the step of decreasing the scores.

4. The method as claimed in claim 2, wherein the step of determining if any of the other active libraries should be deactivated comprises the steps of:

b3a) comparing the scores for the other active libraries to a threshold; and

b3b) if any of the scores is below the threshold, determining that the other active libraries associated with the scores below the threshold should be deactivated.

22. The method as claimed in claim 3, further comprising the steps of:

b2) when the spoken word is in the general library, adjusting scores of other active auxiliary topic libraries;

b3) determining, based on the scores, if any of the other active auxiliary topic libraries should be deactivated; and

,

b4) if any of the other active auxiliary topic libraries should be deactivated, automatically deactivating those of the other active auxiliary topic libraries that should be deactivated.

23. The method as claimed in claim 22, wherein the step of adjusting the scores comprises the step of decreasing the scores.

24. The method as claimed in claim 22, wherein the step of determining if any of the other active auxiliary topic libraries should be deactivated comprises the steps of:

b3a) comparing the scores for the other active auxiliary topic libraries to a threshold; and

b3b) if any of the scores is below the threshold, determining that the other active auxiliary topic libraries associated with the scores below the threshold should be deactivated.

5. A method for dictating speech using a general library and at least one auxiliary topic library, the method comprising the steps of:

a) receiving a spoken word of Input speech;

b) comparing the spoken word to words within one or more active auxiliary topic libraries;

b1) comparing the spoken word to words within a general library;

b2) comparing the spoken word to words within one or more active auxiliary topic libraries;

b3) when the spoken word is recognized in an auxiliary topic library of the one or more active auxiliary topic libraries, adjusting scores of the one or more active auxiliary topic libraries;

b4) determining if any of the one or more active auxiliary topic libraries should be deactivated;

b5) if any of the one or more active auxiliary topic libraries should be deactivated, automatically deactivating those of the one or more active auxiliary topic libraries that should be deactivated;

c) if the spoken word is recognized as being one of the words, dictating the spoken word to be the one of the words and returning to the step of receiving;

d) comparing the spoken word to other words within an inactive auxiliary topic library; and

,

e) if the spoken word is recognized as being one of the other words, automatically activating the inactive auxiliary topic library to be an active auxiliary topic library, dictating the spoken word to be the one of the other words, and returning to the step of receiving.

6. The method as claimed in claim 5, wherein the step of adjusting the scores comprises the steps of:

b3a) increasing a score of the topic library in which the spoken word is recognized; and

b3b) decreasing scores of the one or more active topic libraries in which the spoken word is not recognized.

7. The method as claimed in claim 5, wherein the step of determining if any of the one or more active topic libraries should be deactivated comprises the steps of:

b4a) comparing scores for each of the one or more active topic libraries to a threshold; and

b4b) if any of the scores is below the threshold, determining that the active topic libraries associated with the scores below the threshold should be deactivated.

8. A method for dictating speech using a general library and at least one auxiliary topic library, the method comprising the steps of:

a) receiving a spoken word of Input speech;

b) comparing the spoken word to words within one or more active auxiliary topic libraries;

c) If the spoken word is recognized as being one of the words, dictating the spoken word to be the one of the words and returning to the step of receiving;

d) comparing the spoken word to other words within an Inactive auxiliary topic library;

e) if the spoken word is recognized as being one of the other words, automatically activating the inactive auxiliary topic library to be an active auxiliary topic library, dictating the spoken word to be the one of the other words, and returning to the step of receiving;

f) when the spoken word is recognized as a word In the inactive auxiliary topic library, setting a score associated with the inactive auxiliary topic library to a particular value;

g) adjusting scores of other active auxiliary topic libraries of the one or more active auxiliary topic libraries;

h) determining if any of the other active auxiliary topic libraries should be deactivated; and

i) if any of the other active auxiliary topic libraries should be deactivated, automatically deactivating those of the other active auxiliary topic libraries that should be deactivated.

9. The method as claimed in claim 8, wherein the step of determining if any of the other active topic libraries should be deactivated comprises the steps of:

h1) comparing the scores of each of the other active topic libraries to a threshold; and

h2) if any of the scores is below the threshold, determining that the other active topic libraries associated with the scores below the threshold should be deactivated.

10. A method for dictating speech using a general library and at least one auxiliary topic library, the method comprising the steps of:

a) receiving a spoken word of input speech;

b) comparing the spoken word to words within one or more active auxiliary topic libraries;

c) if the spoken word is recognized as being one of the words, dictating the spoken word to be the one of the words and returning to the step of receiving;

d) comparing the spoken word to other words within an inactive auxiliary topic library;

e) if the spoken word is recognized as being one of the other words, automatically activating the inactive auxiliary topic library to be an active auxiliary topic library, dictating the spoken word to be the one of the other words, and returning to the step of receiving;

f) determining whether a word that has been dictated has been corrected to another word; and

g) if the word that has been dictated has been corrected, determining a library in which the another word exists.

11. The method as claimed in claim 10, further comprising the step of:

h) if the another word exists in a particular inactive library, automatically activating the particular inactive library.

12. The method as claimed in claim 10, further comprising the steps of:

h) adjusting scores of active topic libraries in which the another word does not exist;

i) determining if any of the active topic libraries should be deactivated; and

j) if any of the active topic libraries should be deactivated, automatically deactivating those of the active topic libraries that should be deactivated.

13. An apparatus for dictating speech using a general library and at least one auxiliary topic library comprising;

a microphone that receives input speech;

an analog to digital converter, coupled to the microphone, that converts the input speech Into digital speech samples;

a processor that receives a block of the digital speech samples which represents a spoken word, compares the spoken word to words within one or more active auxiliary topic libraries, compares the spoken word to other words within an inactive auxiliary topic library, and if the spoken word is recognized as being one of the other words, automatically activates the inactive auxiliary topic library to be an active auxiliary topic library, and dictates the spoken word to be the one of the other words; and

a memory device, coupled to the processor, for storing the one or more active auxiliary topic libraries and the inactive auxiliary topic library.

14. A machine readable storage, having stored thereon a computer program having a plurality of code sections executable by a machine for causing the machine to perform the steps of:

a) receiving a spoken word of input speech;

b) comparing the spoken word to words within one or more active auxiliary topic libraries where auxiliary topic libraries are in addition of a general library;

c) if the spoken word is recognized as being one of the words, dictating the spoken word to be the one of the words and returning to the step of receiving;

d) comparing the spoken word to other words within an inactive auxiliary topic library; and

e) if the spoken word is recognized as being one of the other words, automatically activating the inactive auxiliary topic library to be an active auxiliary topic library, dictating the spoken word to be the one of the other words, and returning to the step of receiving.

15. A method for dictating speech using a general library and at least one auxiliary topic library, the method comprising the steps of:

a) receiving a spoken word of input speech;

b) comparing the spoken word to words within one or more active auxiliary topic libraries;

c) if the spoken word is recognized as being one of the words, dictating the spoken word to be the one of the words and returning to the step of receiving;

d) comparing the spoken word to other words within an inactive auxiliary topic library; and

,

e) if the spoken word is recognized as being one of the other words, automatically activating the inactive auxiliary topic library, dictating the spoken word to be the one of the other words, and returning to the step of receiving.

16. The method as claimed in claim 15, further comprising:

f) determining whether an active auxiliary topic library should be deactivated; and

g) if the active auxiliary topic library should be deactivated, automatically deactivating the active auxiliary topic library.

17. The method as claimed in claim 15 wherein the step of comparing the spoken word to words within one or more active libraries includes the step of:

b1) comparing the spoken word to words within a general library.

18. The method as claimed in claim 17, wherein the step of comparing the spoken word to words within one or more active libraries further comprises the step of:

b2) comparing the spoken word to words within one or more active topic libraries.

19. The method as claimed in claim 18, further comprising the steps of:

b3) when the spoken word is recognized in a topic library of the one or more active topic libraries, adjusting scores of the one or more active topic libraries;

b4) determining if any of the one or more active topic libraries should be deactivated; and

b5) if any of the one or more active topic libraries should be deactivated, automatically deactivating those of the one or more active topic libraries that should be deactivated.

20. The method as claimed in claim 19, wherein the step of adjusting the scores comprises the steps of:

b3a) increasing a score of the topic library in which the spoken word is recognized; and

b3b) decreasing scores of the one or more active topic libraries in which the spoken word is not recognized.

21. The method as claimed in claim 19, wherein the step of determining if any of the one or more active topic libraries should be deactivated comprises the steps of:

b4a) comparing scores for each of the one or more active topic libraries to a threshold; and

b4b) if any of the scores is below the threshold, determining that the active topic libraries associated with the scores below the threshold should be deactivated.

25. The method as claimed in claim 15, wherein the step of comparing the spoken word to other words within an inactive topic library comprises the steps of:

d1) comparing the spoken word to words within a quick list which includes a subset of words from the inactive topic library; and

d2) if the spoken word is recognized as being one of the words within the quick list, automatically activating the inactive topic library associated with the quick list.

26. The method as claimed in claim 15, wherein the step of comparing the spoken word to words within one or more active libraries comprises the step of:

b1) determining correlations between the spoken word and active library words; and

b2) determining correlations between the spoken word and inactive library words.

27. The method as claimed in claim 26, further comprising the step of:

f) if one or more of the correlations is greater than a recognition threshold, determining that the spoken word is recognized; and

g) if none of the correlations are greater than the recognition threshold, determining that the spoken word is not recognized.

28. The method as claimed in claim 26, further comprising the steps of:

h) storing correlations associated with closest matching words within each of the one or more active libraries and the inactive library; and

i) if the spoken word is not recognized in the one or more active libraries or the inactive library, determining a word and a library with a highest correlation, and dictating the word with the highest correlation.

29. The method as claimed in claim 15, further comprising the steps of:

f) when the spoken word is recognized as a word in the inactive topic library, setting a score associated with the inactive topic library to a particular value;

g) adjusting scores of other active topic libraries of the one or more active libraries;

h) determining if any of the other active topic libraries should be deactivated;

and

i) if any of the other active topic libraries should be deactivated, automatically deactivating those of the other active topic libraries that should be deactivated.

30. The method as claimed in claim 29, wherein the step of determining if any of the other active topic libraries should be deactivated comprises the steps of:

h1) comparing the scores of each of the other active topic libraries to a threshold; and

h2) if any of the scores is below the threshold, determining that the other active topic libraries associated with the scores below the threshold should be deactivated.

31. The method as claimed in claim 15, further comprising the steps of:

f) determining whether a word that has been dictated has been corrected to another word; and

g) if the word that has been dictated has been corrected, determining a library in which the another word exists.

32. The method as claimed in claim 31, further comprising the step of:

h) if the another word exists in a particular inactive library, automatically activating the particular inactive library.

33. The method as claimed in claim 31, further comprising the steps of:

h) adjusting scores of active topic libraries in which the another word does not exist;

i) determining if any of the active topic libraries should be deactivated; and

j) if any of the active topic libraries should be deactivated, automatically deactivating those of the active topic libraries that should be deactivated.