Script compliance and quality assurance using speech recognition

• US 7,966,187 B1
• Filed: 09/29/2003
• Issued: 06/21/2011
• Est. Priority Date: 02/15/2001
• Status: Expired due to Fees

• Current Assignee: Alorica Business Solutions, LLC
• Original Assignee: West Corporation
• Inventors: Pettay, Mark J., Vacek, Jill M.
• Primary Examiner(s): YEN, ERIC L
• Application Number:
  US10/673,679
• Time in Patent Office: 2,822 Days
• Field of Search: 704270-275, 704/9
• US Class Current: 704/270
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/088 Word spotting

Abstract

Apparatus and methods are provided for using automatic speech recognition to analyze a voice interaction and verify compliance of an agent reading a script to a client during the voice interaction. In one aspect of the invention, a communications system includes a user interface, a communications network, and a call center having an automatic speech recognition component. In other aspects of the invention, a script compliance method includes the steps of conducting a voice interaction between an agent and a client and evaluating the voice interaction with an automatic speech recognition component adapted to analyze the voice interaction and determine whether the agent has adequately followed the script. In yet still further aspects of the invention, a voice record of the given interaction can be analyzed at least in part to determine whether the agent spoke any extraneous material not dictated by the script, apart from or in addition to ensuring that the voice record contains all material required by the script. The voice record can be analyzed against a pre-defined list of specific words or phrases to identify instances of profanity or abusive language used by the agent.

Claims

1. A method for evaluating compliance of at least one agent reading at least one script to at least one client, the method comprising at least the following:

conducting at least one voice interaction between the at least one agent and the at least one client, wherein the at least one agent follows the at least one script at an agent workstation;

evaluating the at least one voice interaction with at least one automatic speech recognition component adapted to analyze the at least one voice interaction, wherein the at least one voice interaction is divided into viewable panel-level segments and a panel-level time displacement stamp is assigned to each panel of the panel-level segments, wherein each panel-level segment is compared with a corresponding expected text, wherein a confidence level threshold of the automatic speech recognition component is used to evaluate the accuracy of each panel-level segment based on an output of a comparison between each panel-level segment and its corresponding expected text, wherein a score is assigned to each panel-level segment, each score indicating a match accuracy between the panel-level segment to which it is assigned and its assigned panel-level segment'"'"'s corresponding expected text, wherein the scores are evaluated against a standard, the standard defining a required score for each of the panel-level segments to be declared as a match to their corresponding expected texts; and

determining whether the at least one agent has adequately followed the at least one script wherein a set of action rules is applied to the output of the determining, wherein the set of action rules includes a quality assurance action to be taken.

2. The script compliance method of claim 1, wherein conducting at least one voice interaction includes conducting at least one voice interaction involving a telemarketing agent.

3. The script compliance method of claim 1, wherein conducting at least one voice interaction includes conducting at least one voice interaction governed by at least one script that includes text corresponding to at least one offer of at least one of goods and services.

4. The script compliance method of claim 1, wherein conducting at least one voice interaction includes conducting the at least one voice interaction at least in part on at least one communications network.

5. The script compliance method of claim 1, wherein conducting at least one voice interaction includes conducting the at least one voice interaction at least in part on a publicly switched telephone network (PSTN).

6. The script compliance method of claim 1, wherein conducting at least one voice interaction includes conducting the at least one voice interaction at least in part on at least one Internet.

7. The script compliance method of claim 1, wherein conducting at least one voice interaction includes conducting the at least one voice interaction at least in part on at least one communications network having at least one wireless component.

8. The script compliance method of claim 1, wherein conducting at least one voice interaction includes conducting at least one telephone call.

9. The script compliance method of claim 1, wherein conducting at least one voice interaction includes conducting at least one telephone call that is initiated by the at least one client.

10. The script compliance method of claim 1, wherein conducting at least one voice interaction includes conducting at least one telephone call that is initiated by an entity other than the at least one client.

11. The script compliance method of claim 1, wherein evaluating the at least one interaction includes at least the following:

converting the at least one voice interaction into at least one digital signal comprising at least one spectral representation of the at least one voice interaction,

comparing the at least one digital signal to at least one reference standard that includes at least one known vocabulary, and

matching the at least one digital signal to at least one of words and phrases contained in the at least one reference standard.

12. The script compliance method of claim 1, further comprising performing at least one action based upon at least one result of the evaluating of the at least one voice interaction.

13. The script compliance method of claim 12, wherein performing at least one action includes transmitting at least one signal to the at least one agent.

14. The script compliance method of claim 12, wherein performing at least one action includes transmitting at least one signal to at least one reviewing authority.

15. The script compliance method of claim 12, wherein performing at least one action includes making at least one entry in at least one script compliance incentive system.

16. The script compliance method of claim 1, further comprising reviewing at least one determination of whether the at least one agent has adequately followed the at least one script.

17. The script compliance method of claim 1, wherein determining whether the at least one agent has adequately followed the at least one script includes defining at least one score assigned by the at least one automatic speech recognition component.

18. The script compliance method of claim 1, wherein evaluating the at least one voice interaction includes evaluating a plurality of panels.

19. The method of claim 1, further comprising comparing data representing an actual duration of at least one interaction, wherein the at least one agent reads at least one script to the at least one client, to data representing an expected duration parameter associated with the at least one interaction.

20. The method of claim 1, further comprising dispositioning at least one interaction, wherein the at least one agent reads at least one script to the at least one client, based at least in part on a comparison of data representing an actual duration of the at least one interaction to data representing an expected duration parameter associated with the at least one interaction.

21. The method of claim 1, wherein determining whether the at least one agent has adequately followed the at least one script includes, at least in part, dispositioning at least one interaction, wherein the at least one agent reads at least one script to the at least one client, based at least in part on a comparison of data representing an actual duration of the at least one interaction to data representing an expected duration parameter associated with the at least one interaction.

22. A system for evaluating compliance of at least one agent reading at least one script to at least one client, the system comprising at least the following:

at least one communication network adapted to support at least one voice interaction between the at least one agent and the at least one client, wherein the at least one agent follows at least one script at an agent workstation;

at least one automatic speech recognition component adapted to analyze the at least one voice interaction and to determine whether the at least one agent has adequately followed the at least one script, wherein the at least one voice interaction is divided into viewable panel-level segments and a panel-level time displacement stamp is assigned to each panel of the panel-level segments, wherein each panel-level segment is compared with a corresponding expected text, wherein a confidence level threshold of the automatic speech recognition component is used to evaluate the accuracy of each panel-level segment based on an output of a comparison between each panel-level segment and its corresponding expected text, wherein a score is assigned to each panel-level segment, each score indicating a match accuracy between the panel-level segment to which it is assigned and its assigned panel-level segment'"'"'s corresponding expected text, wherein the scores are evaluated against a standard, the standard defining a required score for each of the panel-level segments to be declared as a match to their corresponding expected texts; and

at least one set of action rules applied to the output of the determination by the at least one automatic speech recognition component as to whether the at least one agent has adequately followed the at least one script, the at least one set of action rules including a quality assurance action determination.

23. The system of claim 22, wherein the at least one communication network comprises at least one long distance telephone network.

24. The system of claim 22, wherein the at least one communication network comprises at least one internet-based network.

25. The system of claim 22, further comprising at least one call center that includes a plurality of agent workstations.

26. The system of claim 25, wherein at least one of the agent workstations includes at least a telephone and a computer terminal.

27. The system of claim 22, wherein the at least one agent is a telemarketing agent.

28. The system of claim 22, wherein the at least one agent is a customer service agent.

29. The system of claim 22, wherein the means for causing at least one action to be taken includes means for transmitting at least one signal to the at least one agent.

30. The system of claim 22, wherein the means for causing at least one action to be taken includes means for transmitting at least one signal to at least one reviewing authority.

31. The system of claim 22, wherein the means for causing at least one action to be taken includes means for making an entry in a script compliance incentive system.

32. A method of evaluating compliance of at least one agent with at least one script that governs, at least in part, at least one interaction processed by at least one agent, the method comprising at least the following:

creating at least one voice record of at least one interaction processed by the at least one agent at an agent workstation;

defining at least first data relating to evaluating compliance of the at least one agent with the at least one script; and

processing the at least one voice record against the at least first data, wherein the voice record is divided into viewable panel-level segments, wherein a panel-level time displacement stamp is assigned to each panel of the panel-level segments, wherein each panel-level segment is compared with a corresponding portion of the first data, wherein a confidence level threshold of the automatic speech recognition component is used to evaluate the accuracy of each panel-level segment based on an output of a comparison between each panel-level segment and its corresponding portion of the first data, wherein a score is assigned to each panel-level segment, each score indicating a match accuracy between the panel-level segment to which it is assigned and its assigned panel-level segment'"'"'s corresponding portion of the first data, wherein the scores are evaluated against a standard, the standard defining a required score for each of the panel-level segments to be declared as a match to their corresponding portions of the first data, a set of action rules being applied to the output of the processing to direct a quality assurance action to be taken.

33. The method of claim 32, further comprising identifying at least one instance of agent non-compliance with the at least one script based upon the processing of the at least one voice record against the at least first data.

34. The method of claim 33, further comprising processing the at least one instance of agent non-compliance with the at least one script.

35. The method of claim 32, wherein creating at least one voice record includes creating at least one voice record of at least one interaction handled by at least one agent physically located in at least one call center.

36. The method of claim 32, wherein creating at least one voice record includes creating at least one voice record of at least one interaction handled by at least one agent physically located in at least one location remote from at least one call center.

37. The method of claim 32, wherein creating at least one voice record includes creating at least one voice record of at least one interaction handled by at least one remote agent.

38. The method of claim 32, wherein creating at least one voice record includes creating at least one voice record of at least one interaction handled by at least one home-based agent.

39. The method of claim 32, wherein creating at least one voice record includes creating at least one voice record of at least one telephone call received by at least one call center.

40. The method of claim 32, wherein creating at least one voice record includes creating at least one voice record of at least one telephone call originating at least one call center.

41. The method of claim 32, wherein defining at least first data includes defining at least one list containing at least one word appearing in the at least one script.

42. The method of claim 32, wherein defining at least first data includes defining at least one word that is expected to appear in the at least one voice record.

43. The method of claim 32, wherein defining at least first data includes defining at least one list having at least one entry corresponding to at least in part conversion portion of the at least one script.

44. The method of claim 32, wherein defining at least first data includes defining at least one list having at least one entry corresponding to at least one upsell item specified in the at least one script.

45. The method of claim 32, wherein defining at least first data includes defining at least one list containing at least one word not appearing in the at least one script.

46. The method of claim 32, wherein defining at least first data includes defining at least one list containing at least one word that is not expected to appear in the at least one voice record.

47. The method of claim 32, wherein defining at least first data includes defining at least one list containing at least one instance of profane language.

48. The method of claim 32, wherein defining at least first data includes defining at least one list containing at least one instance of abusive language.

49. The method of claim 32, wherein processing the at least one voice record includes evaluating whether the at least one agent spoke at least one word included in the at least one script.

50. The method of claim 32, wherein processing the at least one voice record includes evaluating whether the at least one agent spoke at least one word included in the at least first data.

51. The method of claim 32, wherein processing the at least one voice record includes evaluating whether the at least one agent spoke words that do not appear in the at least one script.

52. The method of claim 51, wherein evaluating whether the at least one agent spoke words include referring to a predefined list of words.

53. The method of claim 51, wherein evaluating whether the at least one agent spoke words includes comparing the voice record to a script governing a given interaction to reveal verbiage spoken by the agent not dictated by the script.

54. The method of claim 32, further comprising identifying at least one instance of agent non-compliance with the at least one script by determining that at least one word contained in the at least one script is not contained in the at least one voice record.

55. The method of claim 32, further comprising identifying at least one instance of agent non-compliance with the at least one script by determining that at least one word contained in the at least first data is not contained in the at least one voice record.

56. The method of claim 32, further comprising identifying at least one instance of agent non-compliance with the at least one script by determining that at least one word contained in the at least first data and not contained in the at least one script appears in the at least one voice record.

57. The method of claim 34, wherein processing the at least one instance of agent non-compliance with the at least one script includes entering data representing the at least one instance of agent non-compliance into a data store.

58. The method of claim 34, wherein processing the at least one instance of agent non-compliance with the at least one script includes entering data representing the at least one instance of agent non-compliance in a data store for subsequent follow-up by at least one supervisor.

59. A program storage device readable by a machine, tangibly embodying a program of instructions executable by the machine to perform a method for evaluating compliance of at least one agent with at least one script that governs, at least in part, at least one interaction processed by at least one agent, the method comprising at least the following:

creating at least one voice record of at least one interaction processed by the at least one agent at an agent workstation;

defining at least first data relating to evaluating compliance of the at least one agent with the at least one script; and

processing the at least one voice record against the at least first data, wherein the voice record is divided into viewable panel-level segments, wherein a panel-level time displacement stamp is assigned to each panel of the panel-level segments, wherein each panel-level segment is compared with a corresponding portion of the first data, wherein a confidence level threshold of the automatic speech recognition component is used to evaluate the accuracy of each panel-level segment based on an output of a comparison between each panel-level segment and its corresponding portion of the first data, wherein a score is assigned to each panel-level segment, each score indicating a match accuracy between the panel-level segment to which it is assigned and its assigned panel-level segment'"'"'s corresponding portion of the first data, wherein the scores are evaluated against a standard, the standard defining a required score for each of the panel-level segments to be declared as a match to their corresponding portions of the first data, a set of action rules being applied to the output of the processing to direct a quality assurance action to be taken.

60. A system for evaluating compliance of at least one agent with at least one script that governs, at least in part, at least one interaction processed by at least one agent, the system comprising at least the following:

means for creating at least one voice record of at least one interaction processed by the at least one agent at an agent workstation;

means for defining at least first data relating to evaluating compliance of the at least one agent with the at least one script; and

means for processing the at least one voice record against a corresponding portion of the at least first data, wherein the voice record is divided into viewable panel-level segments, wherein a panel-level time displacement stamp is assigned to each panel of the panel-level segments, wherein each panel-level segment is compared with the corresponding portion of the first data, wherein a confidence level threshold of the automatic speech recognition component is used to evaluate the accuracy of each panel-level segment based on an output of a comparison between each panel-level segment and its corresponding portion of the first data, wherein a score is assigned to each panel-level segment, each score indicating a match accuracy between the panel-level segment to which it is assigned and its assigned panel-level segment'"'"'s corresponding portion of the first data, wherein the scores are evaluated against a standard, the standard defining a required score for each of the panel-level segments to be declared as a match to their corresponding portions of the first data, a set of action rules being applied to the output of the processing to direct a quality assurance action to be taken.

61. A system for evaluating compliance of at least one agent reading at least one script to at least one client, the system comprising at least the following:

means for conducting at least one voice interaction between the at least one agent and the at least one client, wherein the at least one agent follows the at least one script at an agent workstation;

means for evaluating the at least one voice interaction with at least one automatic speech recognition component adapted to analyze the at least one voice interaction, wherein the at least one voice interaction is divided into viewable panel-level segments and a panel-level time displacement stamp is assigned to each panel of the panel-level segments, wherein each panel-level segment is compared with a corresponding expected text, wherein a confidence level threshold of the automatic speech recognition component is used to evaluate the accuracy of each panel-level segment based on an output of a comparison between each panel-level segment and its corresponding expected text, wherein a score is assigned to each panel-level segment, each score indicating a match accuracy between the panel-level segment to which it is assigned and its assigned panel-level segment'"'"'s corresponding expected text, wherein the scores are evaluated against a standard, the standard defining a required score for each of the panel-level segments to be declared as a match to their corresponding expected texts; and

means for determining whether the at least one agent has adequately followed the at least one script, wherein a set of action rules is applied to the output of the determining, wherein the set of action rules includes a quality assurance action to be taken.

62. A program storage device readable by a machine, tangibly embodying a program of instructions executable by the machine to perform a method for evaluating compliance of at least one agent reading at least one script to at least one client, the method comprising at least the following:

conducting at least one voice interaction between the at least one agent and the at least one client, wherein the at least one agent follows the at least one script at an agent workstation;

evaluating the at least one voice interaction with at least one automatic speech recognition component adapted to analyze the at least one voice interaction, wherein the at least one voice interaction is divided into viewable panel-level segments and a panel-level time displacement stamp is assigned to each panel of the panel-level segments, wherein each panel-level segment is compared with a corresponding expected text, wherein a confidence level threshold of the automatic speech recognition component is used to evaluate the accuracy of each panel-level segment based on an output of a comparison between each panel-level segment and its corresponding expected text, wherein a score is assigned to each panel-level segment, each score indicating a match accuracy between the panel-level segment to which it is assigned and its assigned panel-level segment'"'"'s corresponding expected text, wherein the scores are evaluated against a standard, the standard defining a required score for each of the panel-level segments to be declared as a match to their corresponding expected texts; and

determining whether the at least one agent has adequately followed the at least one script, wherein a set of action rules is applied to the output of the determining, wherein the set of action rules includes a quality assurance action to be taken.