Speech analytics with adaptive filtering

• US 9,257,117 B2
• Filed: 02/04/2014
• Issued: 02/09/2016
• Est. Priority Date: 02/04/2014
• Status: Active Grant

• Current Assignee: Avaya Incorporated
• Original Assignee: Avaya Incorporated
• Inventors: Matula, Valentine C., Shaffer, Shmuel
• Primary Examiner(s): Augustin, Marcellus
• Application Number:
  US14/172,297
• Publication Number:
  US 20150221299A1
• Time in Patent Office: 735 Days
• Field of Search: 704/4, 704/207, 704/211, 704/226, 704/239, 704/246, 704/251
• US Class Current: 1/1
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G10L 25/48 specially adapted for particular use
• CPC Class Codes: G10L 25/51 for comparison or discrimination
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 2201/41 using speaker recognition speaker recognition per se G10L17/00
• CPC Class Codes: H04M 2203/404 Collaboration among agents
• CPC Class Codes: H04M 3/2227 Quality of service monitoring
• 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 3/5175 Call or contact centers supervision arrangements

Abstract

Contact center agents often work in close proximity to other agents. As a primary agent is engaged in a call, a neighboring agent speech may be picked up by the primary agent'"'"'s microphone. Contact centers using automated speech recognition systems may monitor the agent'"'"'s speech for key terms and, if detected, respond accordingly. Determining a primary agent spoke a key term, when the true speaker of the key term is a neighboring agent, may cause errors or other problems. Characterizing at least the primary agent'"'"'s voice and then, once a key term is detected, determining if it was the primary agent that spoke the key term, may help to reduce the errors. Additionally, computational requirements may be reduced as non-key terms may be quickly discarded and optionally, key terms determined to not have been spoke by the primary agent, may also be discarded without further processing.

Claims

1. A method, comprising:

characterizing, by a processor, a number of primary voice attributes of a primary agent;

accessing, by the processor, a voice portion of a communication between the primary agent and a customer on a communications network;

monitoring the voice portion of the communication for a key term;

associating, by the processor, the key term with an expected time context;

determining, by the processor, an actual time context of the key term within the communication;

detecting, by the processor, the key term in the monitored voice portion;

in response to detecting the key term, characterizing a number of term attributes for the key term; and

the processor, upon determining (a) that the number of term attributes for the key term substantially matches the number of primary voice attributes and (b) that the actual time context and the expected time context substantially agree, assigning the key term speaker to be the primary agent.

2. The method of claim 1, further comprising:

the processor, upon determining the number of term attributes for the key term substantially differs from the number of primary voice attributes

assigns the key term speaker to be the primary agent.

3. The method of claim 1, further comprising:

characterizing, by the processor a number of secondary voice attributes of a secondary agent;

determining that the number of term attributes for the key term substantially matches the number of secondary voice attributes; and

upon determining the number of term attributes for the key term substantially matches the number of secondary voice attributes, the processor assigns the key term speaker to be the secondary agent.

4. The method of claim 3, wherein a microphone proximate to the primary agent receives the voice of the secondary agent.

5. The method of claim 3, wherein the primary agent and the secondary agent are engaged in the communication.

6. The method of claim 3, further comprising:

associating, by the processor, the key term with an expected time context;

determining, by the processor, an actual time context of the key term within the communication;

determining, by the processor, the number of term attributes for the key term fails to substantially match the number of primary voice attributes;

determining, by the processor, the number of term attributes for the key term substantially differ from the number of secondary voice attributes; and

upon determining, by the processor, the actual time context and the expected time context substantially do not agree, assigning, by the processor, the key term speaker to be the secondary agent.

7. The method of claim 1, wherein at least one of the term attributes and primary voice attributes comprises at least one of tone, pitch, pace, and volume.

8. The method of claim 1, further comprising:

assigning, by the processor, an action to perform upon the primary agent speaking the key term; and

upon assigning the key term speaker to be the primary agent, performing the action.

9. The method of claim 1, wherein the characterizing of the number of primary voice attributes of the primary agent begins substantially upon the primary agent accepting the communication.

10. A system, comprising:

a processor;

a data repository coupled to the processor; and

a communications network coupled to the processor; and

wherein the processor is programmed to characterize a number of primary voice attributes of a primary agent, access a voice portion of a communication between the primary agent and a customer on the communications network, monitor the voice portion of the communication for a key term, associate the key term with an expected time context, determine an actual time context of the key term within the communication, detect the key term in the monitored voice portion, in response to detecting the key term, characterize a number of term attributes for the key term, and, upon determining (a) that the number of term attributes for the key term substantially matches the number of primary voice attributes and (b)) that the actual time context and the expected time context substantially agree assign the key term speaker to be the primary agent.

11. The system of claim 10, wherein the processor is further operable to:

characterize a number of secondary voice attributes of a secondary agent;

determining that the number of term attributes for the key term substantially matches the number of secondary voice attributes; and

upon determining the number of term attributes for the key term substantially matches the number of secondary voice attributes, assigning the key term speaker to be the secondary agent.

12. The system of claim 11, further comprising:

a microphone proximate to the primary agent and operable to receive the voice of the secondary agent.

13. The system of claim 11, wherein the primary agent and the secondary agent are each nodes on the communication network and the primary agent and the secondary agent are engaged in the communication.

14. A non-transitory memory with instructions stored thereon that when read by a computer cause the computer to perform:

characterizing, by the processor, a number of primary voice attributes of a primary agent;

accessing, by the processor, a voice portion of a communication between the primary agent and a customer;

monitoring, by the processor, the voice portion for a key term;

associating, by the processor, the key term with an expected time context;

determining, by the processor, an actual time context of the key term within the communication;

detecting, by the processor, the key term in the monitored voice portion;

in response to detecting the key term, characterizing a number of term attributes for the key term;

determining, by the processor, that the number of term attributes for the key term substantially matches the number of primary voice attributes; and

the processor, upon determining (a) that the number of term attributes for the key term substantially matches the number of primary voice attributes and (b) that the actual time context and the expected time context substantially agree, assigning the key term speaker to be the primary agent.

15. The instructions of claim 14 being further operable to cause the computer to further perform:

the processor, upon determining the number of term attributes for the key term substantially differs from the number of primary voice attributes

assigning the key term speaker to be the primary agent.

16. The instructions of claim 14 being further operable to cause the computer to further perform:

characterizing, by the processor, a number of secondary voice attributes of a secondary agent;

determining, by the processor, that the number of term attributes for the key term substantially matches the number of secondary voice attributes; and

upon determining the number of term attributes for the key term substantially matches the number of secondary voice attributes, assigning, by the processor, the key term speaker to be the secondary agent.

17. The instructions of claim 16, wherein a microphone proximate to the primary agent receives the voice of the secondary agent.

18. The instructions of claim 16, wherein the primary agent and the secondary agent are engaged in the communication.

19. The instructions of claim 16 being further operable to cause the computer to further perform:

associating, by the processor, the key term with an expected time context;

determining, by the processor, an actual time context of the key term within the communication;

determining, by the processor, the number of term attributes for the key term fails to substantially match the number of primary voice attributes;

determining, by the processor, the number of term attributes for the key term substantially differ from the number of secondary voice attributes; and

upon determining the actual time context and the expected time context substantially do not agree, assigning, by the processor, the key term speaker to be the secondary agent.

20. The instructions of claim 14 being further operable to cause the computer to further perform:

assigning, by the processor, an action to perform upon the primary agent speaking the key term; and

upon assigning the key term speaker to be the primary agent, performing the action.