Method and device for automatic recognition of given keywords and/or terms within voice data

• US 9,064,494 B2
• Filed: 09/13/2010
• Issued: 06/23/2015
• Est. Priority Date: 09/11/2009
• Status: Active Grant

• Current Assignee: Vodafone GmbH (Vodafone Group PLC)
• Original Assignee: Vodafone GmbH (Vodafone Group PLC)
• Inventors: Holtel, Stefan
• Primary Examiner(s): YEN, ERIC L
• Application Number:
  US12/807,750
• Publication Number:
  US 20110125501A1
• Time in Patent Office: 1,744 Days
• Field of Search: 704270-275
• US Class Current: 1/1
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G10L 2015/221 Announcement of recognition results
• CPC Class Codes: H04M 3/567 Multimedia conference systems

Abstract

The present invention relates to a method of and a device (30) for automatic recognition of given keywords and/or terms within voice data (25) of a talk between at least two participants, said voice data (25) being continuously compared during said talk with said given keywords and/or terms (26) with regard to the occurrence of said given keywords and/or terms (26) within said voice data (25). In order to provide a solution which guarantees that topics which should be part of a talk are actually dealt with during said talk, the method is characterized in that a visualized representation (48) of the results (45) of said comparison is presented to at least one participant during the talk. A device (30) for automatic recognition of given keywords and/or terms (26) within voice data (25) of a talk between at least two participants comprises a comparator unit (41) for continuously comparing said voice data (25) with said given keywords and/or terms (25) during said talk with regard to the occurrence of said given keywords and/or terms (26) within said voice data (25); and means (46) for transforming the results (45) of said comparison into a visualized representation (48) and for presenting said visualized representation (48) of the comparison results (45) to at least one participant during the talk.

Claims

1. A method of automatic recognition of given keywords and/or terms within voice data of a talk between at least two participants, said voice data being continuously electronically compared during said talk with said given keywords and/or terms with regard to the occurrence of said given keywords and/or terms within said voice data,

characterized

in that during said comparison step it is checked, whether a given keyword and/or term occurs during said talk and, if the given keyword and/or term occurs during said talk, how often the given keyword and/or term occurs during said talk,

in that all given keywords and/or terms are displayed as a visualized representation at the beginning of the talk, and

in that if one of the keywords and/or terms is used during the talk or if the frequency of one of the given keywords and/or terms reaches a predetermined threshold, said one of the keywords and/or terms is removed from the visualized representation.

2. The method according to claim 1, characterized in that the given keywords and/or terms are talk type related keywords and/or terms.

3. The method according to claim 1, characterized in that the given keywords and/or terms are electronically stored within a data file, that the comparison of the voice data with the given keywords and/or terms is executed within a processing unit and that the processing unit accesses said data file at least temporarily.

4. The method according to claim 3, characterized in that the data file of given keywords and/or terms is generated by at least one participant or that the data file of given keywords and/or terms is provided as a predefined data file.

5. The method according to claim 1, characterized in that the comparison of the voice data with the given keywords and/or terms is executed within a processing unit, that the voice data are converted into text data by use of a means for automatic voice recognition and that the processing unit accesses said text data, that the processing unit further accesses a data file of given keywords and/or terms, that the given keywords and/or terms are compared with said text data within said processing unit, that matching keywords and/or terms are extracted from said text data, and that said extracted matching keywords and/or terms are displayed as a visualized representation.

6. The method according to claim 1, characterized in that a weighting procedure is electronically performed for those given keywords and/or terms which occur within the voice data, said weighting procedure being performed with regard to the frequency of said given keywords and/or terms within said voice data, and that the given keywords and/or terms which occur within said voice data, are displayed as a visualized representation according to their weight.

7. The method according to claim 1, characterized in that the visualized representation of the comparison results is graphically presented by a diagram, or by a call tag cloud, or by deleting at least parts of said comparison results.

8. The method according to claim 1, characterized in that a communication terminal is allocated to each participant of the talk, that during said talk voice data are exchanged between said communication terminals and that a visualized representation of the comparison results is displayed on at least one communication terminal of at least one participant.

9. The method according to claim 1, characterized in that the method is executed within a processing unit, said processing unit being provided on the operator side of a communications network.

10. The method according to claim 1, characterized in that the method is executed within a processing unit, said processing unit being part of a communication terminal being registered with a communications network.

12. The device according to claim 11, characterized by means for performing the method according to claim 1.

11. A device for automatic recognition of given keywords and/or terms within voice data of a talk between at least two participants, said communication device comprising:

a comparator unit for continuously comparing said voice data with said given keywords and/or terms during said talk with regard to the occurrence of said given keywords and/or terms within said voice data, said comparator unit being adapted for checking, whether a given keyword and/or term occurs during said talk and, if the given keyword and/or term occurs during said talk, how often the given keyword and/or term occurs during said talk; and

means for presenting, to at least one participant, a visualized representation of keywords and/or terms, the visualized representation including, at the beginning of the talk, all given keywords and/or terms and being modified during the talk to reflect comparison results so that if one of the given keywords and/or terms is used during the talk or if the frequency of one of the given keywords and/or terms reaches a predetermined threshold, said one of the keywords and/or terms is removed from the visualized representation.

13. The device according to claim 11, characterized in that it comprises means for automatic voice recognition for converting the voice data into text data.

14. The device according to claim 11, characterized in that the device is at least one component on the operator side of a communications network.

15. The device according to claim 11, characterized in that the device is at least one component of a communication terminal being registered with a communications network.