System and method for extracting data from audio messages

• US 6,249,765 B1
• Filed: 12/22/1998
• Issued: 06/19/2001
• Est. Priority Date: 12/22/1998
• Status: Expired due to Term

• Current Assignee: Majandro LLC (IP Edge LLC)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Adler, Annette M., Fishkin, Kenneth P., Marshall, Catherine C., Silverman, Alexander E., Cass, Todd A.
• Primary Examiner(s): Dorvil, Richemond
• Assistant Examiner(s): Nolan, Daniel A.
• Application Number:
  US09/219,091
• Time in Patent Office: 910 Days
• Field of Search: 704/270-272, 704/231
• US Class Current: 704/500
• CPC Class Codes: G10L 15/04 Segmentation; Word boundary detection
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G11B 27/034 on discs G11B27/036, G11B27/038 take precedence
• CPC Class Codes: G11B 27/28 by using information signals recorded by the same method as the main recording G11B27/22 takes precedence
• CPC Class Codes: G11B 27/34 Indicating arrangements indicating means incorporated in magazine or cassette G11B23/046 and G11B23/0875; indicating measured values in general G01D
• CPC Class Codes: H04M 1/57 Arrangements for indicating or recording the number of the calling subscriber at the called subscriber's set at the operator set in a manual exchange H04M5/20

Abstract

A system and method for extracting key information from digitized audio messages, including telephone voice messages. Information, such as a telephone number and the name of the caller, is derived and extracted from a voice message and used to establish links to the information within the message. The telephone number and name of the caller can then be replayed without the need to replay the entire voice message. The telephone number and name of the caller can also be used as indices into an information database.

Claims

1. A method for manipulating and managing information in a digitized data stream, comprising the steps of:

receiving the digitized data stream, wherein the digitized data stream comprises a digitized audio message;

receiving guidance representative of the structure of the digitized data stream;

analyzing the digitized data stream, wherein the analyzing step comprises performing a speech recognition operation on the digitized audio message, identifying a plurality of words within the digitized audio message, and storing the position within the digitized audio message of each word;

determining whether a portion of the audio message represents a telephone number or a name;

using the guidance to identify at least one significant data item within the digitized data stream and selecting a date contained within the digitized data stream as the data item; and

storing a pointer to the significant data item.

2. The method of claim 1, wherein the speech recognition operation utilizes a Hidden Markov Model.

3. The method of claim 1, wherein the speech recognition operation has a general-purpose vocabulary.

4. The method of claim 1, wherein the speech recognition operation has a limited vocabulary, comprising the numerals “

zero”

through “

nine” and

at least one cue word.

5. The method of claim 1, wherein the analyzing step comprises the steps of:

identifying at least one pause in the digitized audio message;

identifying at least one cue word in the digitized audio message; and

storing the position within the digitized audio message of each pause and cue word.

6. The method of claim 1, wherein the analyzing step comprises the step of receiving at least one manual cue representative of a position in the digitized audio message.

7. The method of claim 1, wherein the step of determining whether a portion of the audio message represents a telephone number is responsive to user input.

8. The method of claim 1, wherein the step of determining whether a portion of the audio message represents a name is responsive to user input.

9. The method of claim 1, further comprising the step of enabling playback of the data stream starting at the pointer.

10. The method of claim 1, wherein the guidance comprises user input.

11. The method of claim 1, wherein the guidance comprises a template for the digitized data stream.

12. The method of claim 1, wherein the guidance comprises a genre specification for digitized data stream.