Multi-pass speech analytics

• US 9,171,547 B2
• Filed: 10/12/2011
• Issued: 10/27/2015
• Est. Priority Date: 09/29/2006
• Status: Active Grant

• Current Assignee: Verint Americas Incorporated (Verint Systems Incorporated)
• Original Assignee: Verint Americas Incorporated (Verint Systems Incorporated)
• Inventors: Watson, Joseph
• Primary Examiner(s): Han, Qi
• Application Number:
  US13/271,681
• Publication Number:
  US 20120026280A1
• Time in Patent Office: 1,476 Days
• Field of Search: 704/251, 704/270, 704/246, 704/250, 704/231, 704/235, 704/273
• US Class Current: 1/1
• CPC Class Codes: G10L 15/32 Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems

Abstract

Included are embodiments for multi-pass analytics. At least one embodiment of a method includes receiving data associated with a communication, performing first tier analytics on the received data, and performing second tier analytics on the received data, where the second tier analytics determines different characteristics of the received data than the first tier analytics.

Claims

1. A method of performing data analytics in a computing system, the method comprising:

receiving audio data associated with a communication from a caller to be analyzed;

performing a first analytics process by a first tier speech recognition server, the first analytics process comprising;

converting the received audio data into a phonetic transcript,

determining at least one characteristic associated with the audio data to determine whether speaker identification can be determined by a second tier speech recognition server, wherein the second tier speech recognition server is configured as a speaker identification determination server, and wherein the first tier recognition server is separately located from the second tier speech recognition server, and

if the determination is made that speaker identification can be determined by the second tier speech recognition server, sending first data comprising a portion of the phonetic transcript to the second tier recognition server; and

performing a second analytics process on the first data by the second tier speech recognition server, different than the first analytics process, the second analytics process comprising analyzing the first data to determine the identification of the speaker.

2. The method of claim 1, further comprising providing the first data to the second analytics process to perform analytics in accordance with information contained in the first data.

3. The method of claim 2, further comprising:

determining a subset of the first data in accordance with an event; and

performing analytics using the subset of the first data.

4. The method of claim 2, further comprising performing speech recognition using the first data.

5. The method of claim 1, wherein the first analytics process and the second analytics process determine different characteristics of the speech-to-text converted communication.

6. A non-transitory computer readable medium having stored thereon instructions that, when executed by a processor, direct the processor to:

receive audio data associated with a communication from a caller to be analyzed;

perform a first analytics process by a first tier speech recognition server, the first analytics processes comprising;

converting the received audio data into a phonetic transcript,

determining at least one characteristic associated with the audio data to determine whether speaker identification can be determined by a second tier speech recognition server, wherein the second tier speech recognition server is configured as a speaker identification determination server, and wherein the first tier recognition server is separately located from the second tier speech recognition server, and

if the determination is made that speaker identification can be determined by the second tier speech recognition server, sending first data comprising a portion of the phonetic text to the second tier speech recognition server; and

perform a second analytics process on the first data by the second tier recognition server, different than the first analytics process, the second analytics process comprising analyzing the first data to determine the identification of the speaker.