MASS ELECTRONIC QUESTION FILTERING AND ENHANCEMENT SYSTEM FOR AUDIO BROADCASTS AND VOICE CONFERENCES

• US 20100076747A1
• Filed: 09/25/2008
• Published: 03/25/2010
• Est. Priority Date: 09/25/2008
• Status: Abandoned Application

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Mathai, Shiju, Weisbard, Keeley L., Appleyard, James P.
• Application Number:
  US12/238,246
• Publication Number:
  US 20100076747A1
• Time in Patent Office: Days
• US Class Current: 704/3
• CPC Class Codes: G10L 15/005 Language recognition
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A system for providing electronic filtering and enhancement for audio broadcasts and voice conferences. The system can comprise one or more computing devices configured to record one or more spoken segments, wherein the one or more spoken segments are comprised of utterances. The system can also include one or more electronic data processors configured to process, manage, and store the one or more spoken segments and data, wherein the at least one electronic data processor is communicatively linked to the one or more computing devices. The system can further include a speech-to-text module configured to execute on the one or more electronic data processors, wherein the speech-to-text module converts the one or more spoken segments into a plurality of text segments. Additionally, the system can include a database module configured to execute on the one or more electronic data processors, wherein the database module stores the plurality of text segments in a queue. The system can also include a filtration-prioritization module configured to execute on the one or more electronic data processors, wherein the filtration-prioritization module is configured to filter one or more text segments of the plurality of text segments in the queue, wherein the utterances to be filtered arc defined in advance of filtering. The filtration-prioritization module can also be configured to determine a relevance of the one or more text segments. The filtration-prioritization module can be further configured to prioritize the one or more text segments based upon one or more of the relevance and a similarity of the one or more text segments to other text segments of the plurality of text segments in the queue. Moreover, the filtration-prioritization module can be configured to transmit the one or more text segments to a presenter.

Claims

1. A computer-based method for providing electronic filtering and enhancement in a system for audio broadcasts and voice conferences the method comprising:

recording at least one spoken segment, wherein the at least one spoken segment is comprised of utterances;

converting the at least one spoken segment into a plurality of text segments and storing the plurality of text segments in a queue;

filtering at least one text segment of the plurality of text segments in the queue wherein the utterances to be filtered are defined in advance of filtering;

prioritizing the at least one text segment based upon at least one of a relevance of the at least one text segment and a similarity of the at least one text segment to other text segments of the plurality of text segments in the queue; and

transmitting the at least one text segment to a presenter.

2. The method of claim 1, wherein the at least one spoken segment is associated with a topic of the presenter and further comprising determining the relevance based on a correlation of the at least one text segment with the topic.

3. The method of claim 1, wherein the recording of the at least one spoken segment can be initiated by pressing a key on a device and terminated by pressing the key again and wherein the at least one spoken segment is disassociated from a particular user making the at least one spoken segment.

4. The method of claim 1, further comprising determining a language of the presenter.

5. The method of claim 4, further comprising analyzing the at least one spoken segment to determine if the at least one spoken segment is in the determined language of the presenter.

6. The method of claim 5, further comprising translating the at least one spoken segment to the determined language of the presenter if the at least one spoken segment is determined to be in a language different from the determined language of the presenter.

7. The method of claim 1, wherein the filtering step comprises excluding other text segments of the plurality of text segments similar to the at least one text segment in the queue.

8. The method of claim 7, further comprising adding a bonus score to the at least one text segment which had similar other text segments excluded, wherein the bonus score corresponds to the quantity of similar other text segments excluded and enables the at least one text segment to be prioritized higher in the queue.

9. The method of claim 1, further comprising filtering the at least one text segment using a keyword, wherein the keyword is matched to an utterance contained within the at least one text segment and can be used to perform at least one of excluding, including, and prioritizing the at least one text segment.

10. The method of claim 9, wherein the keyword can be assigned a weight, wherein the weight indicates the relevance of the particular keyword.

11. The method of claim 1, wherein the filtering and prioritizing steps are performed by a moderator.

12. The method of claim 11, wherein the moderator can edit the at least one text segment and deliver the at least one text segment to the presenter.

13. A computer-based system for providing electronic filtering and enhancement for audio broadcasts and voice conferences, the system comprising:

at least one computing device configured to record at least one spoken segment, wherein the at least one spoken segment is comprised of utterances;

at least one electronic data processor configured to process, manage, and store the at least one spoken segment and data, wherein the at least one electronic data processor is communicatively linked to the at least one computing device;

a speech-to-text module configured to execute on the at least one electronic data processor, wherein the speech-to-text module converts the at least one spoken segment into a plurality of text segments;

a database module configured to execute on the at least one electronic data processor, wherein the database module stores the plurality of text segments in a queue;

a filtration-prioritization module configured to execute on the at least one electronic data processor, wherein the filtration-prioritization module is configured to;

filter at least one text segment of the plurality of text segments in the queue, wherein the utterances to be filtered are defined in advance of filtering;

determine a relevance of the at least one text segment;

prioritize the at least one text segment based upon at least one of the relevance and a similarity of the at least one text segment to other text segments of the plurality of text segments in the queue; and

transmit the at least one text segment to a presenter.

14. The system of claim 13, wherein the at least one spoken segment is associated with a topic of the presenter and further comprising determining the relevance based on a correlation of the at least one text segment with the topic.

15. The system of claim 13, wherein the recording of the at least one spoken segment can be initiated by pressing a key on the at least one computing device and terminated by pressing the key again and wherein the at least one spoken segment can be disassociated from a particular user making the at least one spoken segment.

16. The system of claim 13, further comprising a language analyzer module configured to execute on the at least one electronic data processor, wherein the language analyzer module is configured to determine a language of the presenter.

17. The system of claim 16, wherein the language analyzer module is further configured to analyze the at least one spoken segment to determine if the at least one spoken segment is in the determined language of the presenter.

18. The system of claim 17, further comprising a language translator module configured to execute on the at least one electronic data processor, wherein the language translator module is configured to translate the at least one spoken segment to the determined language of the presenter if the at least one spoken segment is determined to be in a language different from the determined language of the presenter.

19. The system of claim 13, wherein the filtration-prioritization module excludes other text segments of the plurality of text segments similar to the at least one text segment in the queue.

20. The system of claim 19, further comprising adding a bonus score to the at least one text segment which had similar other text segments excluded, wherein the bonus score corresponds to the quantity of similar other text segments excluded and enables the at least one text segment to be prioritized higher in the queue.

21. The system of claim 13, wherein the filtration-prioritization module filters the at least one text segment using a keyword, wherein the keyword is matched to an utterance contained within the at least one text segment and can be used to perform at least one of excluding, including, and prioritizing the at least one text segment.

22. The system of claim 21, wherein the keyword can be assigned a weight, wherein the weight indicates the relevance of the particular keyword.

23. The system of claim 13, wherein the filtering and prioritizing are performed by a moderator.

24. The system of claim 23, wherein the moderator can edit the at least one text segment and deliver the at least one text segment to the presenter.

25. A computer-readable storage medium having stored therein computer-readable instructions, which, when loaded in and executed by a computer causes the computer to perform the steps of:

recording at least one spoken segment, wherein the at least one spoken segment is comprised of utterances;

converting the at least one spoken segment into a plurality of text segments and storing the plurality of text segments in a queue;

filtering at least one text segment of the plurality of text segments in the queue, wherein the utterances to be filtered are defined in advance of filtering;

determining a relevance of the at least one text segment;

determining a similarity of the at least one text segment to other text segments of the plurality of text segments in the queue;

prioritizing the at least one text segment based upon at least one of the determined relevance and the determined similarity; and

transmitting the at least one text segment to a presenter.

26. The computer-readable storage medium of claim 25, wherein the at least one spoken segment is associated with a topic of the presenter and further comprising determining the relevance based on a correlation of the at least one text segment with the topic.

27. The computer-readable storage medium of claim 25, wherein the recording step, the recording of the at least one spoken segment can be initiated by pressing a key on a device and terminated by pressing the key again and wherein the at least one spoken segment can be disassociated from a particular user making the at least one spoken segment.

28. The computer-readable storage medium of claim 25, further comprising computer-readable code for causing the computer to determine a language of the presenter.

29. The computer-readable storage medium of claim 28, further comprising computer-readable code for causing the computer to analyze the at least one spoken segment to determine if the at least one spoken segment is in the determined language of the presenter.

30. The computer-readable storage medium of claim 29, further comprising computer-readable code for causing the computer to translate the at least one spoken segment to the determined language of the presenter if the at least one spoken segment is determined to be in a language different from the determined language of the presenter.

31. The computer-readable storage medium of claim 25, wherein the filtering step comprises excluding other text segments of the plurality of text segments similar to the at least one text segment in the queue.

32. The computer-readable storage medium of claim 31, further comprising computer-readable code for causing the computer to add a bonus score to the at least one text segment which had similar other text segments excluded, wherein the bonus score corresponds to the quantity of similar other text segments excluded and enables the at least one text segment to be prioritized higher in the queue.

33. The computer-readable storage medium of claim 25, wherein the filtering step comprises filtering the at least one text segment using a keyword, wherein the keyword is matched to an utterance contained within the at least one text segment and can be used to perform at least one of excluding, including, and prioritizing the at least one text segment.

34. The computer-readable storage medium of claim 33, wherein the keyword can be assigned a weight, wherein the weight indicates the relevance of the particular keyword.

35. The computer-readable storage medium of claim 25, wherein the filtering and prioritizing steps are performed by a moderator and wherein the moderator can edit the at least one text segment and deliver the at least one text segment to the presenter.