Contact center speech analytics system having multiple speech analytics engines

• US 9,014,364 B1
• Filed: 03/31/2014
• Issued: 04/21/2015
• Est. Priority Date: 03/31/2014
• Status: Active Grant

• Current Assignee: Alvaria Cayman
• Original Assignee: Noble Systems Corporation
• Inventors: Koster, Karl H., Haggerty, Christopher S.
• Primary Examiner(s): Nguyen, Khai N
• Application Number:
  US14/230,328
• Time in Patent Office: 386 Days
• Field of Search: 379/265.01, 379/265.02, 379/265.03, 379/265.04, 379/265.05, 379/265.06, 379/265.07, 379/265.08, 379/265.09, 379/265.1, 379/265.11, 379/265.12, 379/265.13, 379/265.14, 379/266.01, 379/266.02, 379/266.03, 379/266.04, 379/266.05, 379/266.06, 379/266.07, 379/266.08, 379/266.09, 379/266.1
• US Class Current: 379/265.07
• 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 2015/088 Word spotting
• CPC Class Codes: G10L 25/63 for estimating an emotional state
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/42059 Making use of the calling party identifier
• CPC Class Codes: H04M 3/42221 Conversation recording systems at the subscriber's set H04M1/656
• CPC Class Codes: H04M 3/4936 Speech interaction details speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/51 Centralised call answering arrangements requiring operator intervention , e.g. call or contact centers for telemarketing
• CPC Class Codes: H04M 3/5166 in combination with interactive voice response systems or voice portals, e.g. as front-ends
• CPC Class Codes: H04M 3/5175 Call or contact centers supervision arrangements

Abstract

Various embodiments of the invention provide methods, systems, and computer-program products for providing a plurality of speech analytics engines in a speech analytics module for detecting semantic and non-semantic speech characteristics in the audio of a call involving an agent in a contact center and a remote party. The speech analytics module generates event notification messages reporting the detected semantic and non-semantic speech characteristics and these messages are sent to an event handler module that forwards the messages to one or more application specific modules. In turn, the application specific modules provide functionality based on the semantic and non-semantic speech characteristics detected during the call such as, for example, causing information to be presented on the screen of a computer used by the agent during the call.

Claims

1. A method for providing speech-related characteristics of speech audio during a call involving a remote party and an agent of a call center, comprising:

bridging in a speech analytics system (“

SAS”

) to the call, wherein the SAS receives the speech audio from at least one of the agent and the remote party;

analyzing the speech audio by the SAS to ascertain a semantic speech characteristic of the speech audio using a semantic speech analytics engine, the semantic speech characteristic indicating a meaning of speech or uttered word made by at least one of the remote party and the agent;

analyzing the speech audio by the SAS to ascertain a non-semantic speech characteristic of the speech audio using a non-semantic speech analytics engine, the non-semantic speech characteristic being pertinent to at least one of the remote party and the agent or delivery of the speech by at least one of the remote party and the agent that does not directly indicate the meaning of the speech or the uttered word made by at least one of the remote party and the agent;

determining by the SAS to send the semantic speech characteristic to a first application specific module (“

ASM”

) and not to a second ASM, the first ASM configured to perform functionality for a first particular application and the second ASM configured to perform functionality for a second particular application in which the first particular application is different than the second particular application;

determining by the SAS to send the non-semantic speech characteristic to the second ASM and not to the first ASM;

providing the semantic speech characteristic using a first speech-related event notification message to the first ASM and the non-semantic speech characteristic using a second speech-related event notification message to the second ASM;

performing the functionality for the first particular application by the first ASM in response to receiving the first speech-related event notification message; and

performing the functionality for the second particular application by the second ASM in response to receiving the second speech-related event notification message.

2. The method of claim 1 further comprising:

generating a call instance data structure in the first ASM in response to receiving a call-related event notification message prior to receiving the first speech-related event notification message, wherein the call instance data structure stores an identifier of the call and the first speech-related event notification message comprises the identifier of the call.

3. The method of claim 1, wherein performing the first particular application by the first ASM comprises:

causing a window to be displayed on a computer used by the agent; and

performing the second particular application by the second ASM comprises;

causing a supervisor to bridge onto the call.

4. The method of claim 1 further comprising:

providing the speech audio in parallel to the semantic speech analytics engine and the non-semantic speech analytics engine.

5. A speech analytics system for use in a contact center comprising:

one or more computer processors; and

memory storing a speech analytics module (SAM), a first application specific module (ASM) configured to perform functionality for a first particular application, and a second configured to perform functionality for a second particular application that is different from the first particular application, wherein

the one or more computer processors execute the SAM to;

receive speech audio of a call involving an agent and a remote party,

analyze the speech audio to detect a semantic speech characteristic and a non-semantic speech characteristic in the speech audio, the semantic speech characteristic indicating a meaning of speech or uttered word made by at least one of the agent and the remote party and the non-semantic speech characteristic being pertinent to at least one of the agent and the remote party or delivery of the speech by at least one of the agent and the remote party that does not directly indicate the meaning of the speech or the uttered word made by at least one of the agent and the remote party,

determine to send the semantic speech characteristic to the first ASM and not to the second ASM,

determine to send the non-semantic speech characteristic to the second ASM and not to the first ASM,

send a first speech-related event notification message that identifies the semantic speech characteristic to the first ASM, and

send a second speech-related event notification message that identifies the non-semantic speech characteristic to the second ASM;

the one or more computer processors execute the first ASM to;

receive the first speech-related event notification message, and

perform the functionality for the first particular application in response to receiving the first speech-related event notification message; and

the one or more computer processors execute the second ASM to;

receive the second speech-related event notification message, and

perform the functionality for the second particular application in response to receiving the second speech-related event notification message.

6. The system of claim 5, wherein the first particular application comprises causing data to display on a pop-up window on a computer used by the agent and second particular application comprises causing a supervisor to bridge onto the call.

7. The system of claim 5, wherein the one or more computer processors execute the first ASM to:

generate a call instance data structure in response to receiving a call-related event notification message prior to receiving the first speech-related event notification message, the call instance data structure configured to store an identifier of the call and the first speech-related event notification message comprises the identifier of the call.

8. The system of claim 5, wherein the one or more computer processors execute the SAM to generate the first speech-related event notification message and the second speech-related event notification message with a call reference identifying the call.

9. A non-transitory, computer-readable medium comprising computer-executable instructions that when executed cause at least one computer processor to:

receive a call-related event notification message indicating an existence of a call, the call-related event notification message comprising a call reference associated with the call;

generate a call instance data structure associated with the call reference in response to receiving the call-related event notification message;

receive at least one speech-related event notification message originating from a speech analytics module, the at least one speech-related event notification message comprising a first indicator of a semantic speech characteristic detected in speech audio of the call indicating a meaning of speech or uttered word made by at least one party on the call, a second indicator of a non-semantic speech characteristic detected in the speech audio of the call being pertinent to at least one party on the call or delivery of the speech by at least one party on the call that does not directly indicate the meaning of the speech or the uttered word made by at least one party on the call, and the call reference;

in response to receiving the at least one speech-related event notification message, perform a first particular application; and

store data indicating receiving of the at least one speech-related event notification message in the call instance data structure.

10. The non-transitory, computer-readable medium of claim 9, wherein the instructions when executed further cause the at least one computer processor to:

receive a second call-related event notification message indicating a termination of the call, the second call-related event notification message comprising the call reference; and

store at least a portion of the data in the call instance data structure in a non-volatile memory.

11. A non-transitory, computer-readable medium comprising computer-executable instructions that when executed cause at least one computer processor to:

receive an indication of an establishment of a call between an agent and a remote party;

receive speech audio associated with the call;

ascertain a plurality of speech analytics engines to analyze the speech audio, wherein the plurality of speech analytics engines comprises a first speech analytics engine to process the speech audio to ascertain semantic aspects of the speech audio indicating a meaning of speech or uttered word made by at least one of the agent and the remote party and a second speech analytics engine to process the speech audio to ascertain non-semantic aspects of the speech audio being pertinent to at least one of the agent and the remote party or delivery of the speech by at least one of the agent and the remote party that does not directly indicate the meaning of the speech or the uttered word made by at least one of the agent and the remote party; and

provide the speech audio to the plurality of speech analytics engines for the duration of the call;

determine to transmit an event notification message to a first application specific module based on the second speech analytics engine ascertaining a non-semantic aspect of the speech audio, the first application specific module configured to perform functionality for a first particular application;

determine not to transmit the event notification message to a second application specific module, the second application specific module configured to perform functionality for a second particular application that is different from the first particular application; and

transmit the event notification message to the first application specific module.

12. The non-transitory, computer-readable medium of claim 11, wherein the instructions when executed further cause the at least one computer processor to:

provide the speech audio in a parallel manner to the plurality of speech analytics engines.

13. The non-transitory, computer-readable medium storing instructions of claim 11, wherein the instructions when executed further cause the at least one computer processor to:

provide the speech audio to the first speech analytics engine and then to the second speech analytics engine after providing the speech audio to the first speech analytics engine.

14. The non-transitory, computer readable medium of claim 11, wherein the instructions when executed further cause the at least one computer processor to:

receive an occurrence message generated by the second speech analytics engine, the occurrence message identifying the non-semantic aspect of the speech audio and having a first protocol structure; and

generate the event notification message having a second protocol structure based on the occurrence message generated by the second speech analytics engine.