System, method, and computer program for speech recognition assisted call center and self service interface

• US 9,191,476 B1
• Filed: 01/08/2009
• Issued: 11/17/2015
• Est. Priority Date: 01/08/2009
• Status: Active Grant

• Current Assignee: Amdocs Development Limited (Amdocs Ltd.), Amdocs Software Systems Limited (Amdocs Ltd.)
• Original Assignee: Amdocs Software Systems Limited (Amdocs Ltd.)
• Inventors: Givoly, Tal
• Primary Examiner(s): Bezuayehu, Solomon
• Application Number:
  US12/350,880
• Time in Patent Office: 2,504 Days
• Field of Search: 37926501-26601, 379 8801- 8819, 455/414.1
• US Class Current: 1/1
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces
• CPC Class Codes: H04M 1/271 controlled by voice recognition
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 2203/6054 Biometric subscriber identification
• CPC Class Codes: H04M 3/42 Systems providing special services or facilities to subscribers specially adapted for wireless communication networks H04W4/00
• CPC Class Codes: H04M 3/5183 Call or contact centers with computer-telephony arrangements

Abstract

A system, method, and computer program are provided for using speech recognition to assist call center interactions with a caller. In operation, utterances are received between a caller and an agent. Additionally, speech recognition is performed on the utterances utilizing a context of the utterances. Furthermore, an interaction with the caller is assisted utilizing the speech recognition.

Claims

1. A method, comprising:

sending, via an interface, from an agent to a caller, a request for authentication information for the caller;

receiving, in response to the request for authentication information, a plurality of utterances from the caller, the plurality of utterances including a plurality of spoken words;

performing, by at least one processor, speech recognition on the plurality of spoken words to generate text representing the utterances;

identifying, by the at least one processor, a context of the utterances, using the text, wherein the context including a particular pattern of the utterances and a particular grammar used in the utterances;

determining, by the at least one processor, a problem and location information associated with the caller using the identified context;

in response to determining that the caller has not provided expected authentication information in the identified context, determining, by at least one processor, whether a solution to the problem is available based on the location information;

when it is determined that the solution to the problem is available, providing, by the at least one processor, the solution to the caller, via the interface, without authenticating the caller;

when it is determined that the solution to the problem is not available, requesting, via the interface, additional authentication information from the caller; and

verifying, by the at least processor, text entered by the agent matches the recognized spoken words; and

in response to detecting a mismatch, alerting the agent, via agent interface, that the mismatch has been detected.

2. The method of claim 1, wherein the utterances include phrases.

3. The method of claim 2, wherein the phrases are parsed to identify parsed text within the phrases, such that fields of a database are filled with the parsed text.

4. The method of claim 3, wherein the parsed text is representative of only a subset of the utterances.

5. The method of claim 1, wherein the context includes at least one semantic aspect associated with the utterances.

6. The method of claim 5, wherein the performing is carried out as a function of the semantic aspect.

7. The method of claim 1, wherein the speech recognition includes converting the utterances to the text and the text is marked in response to the conversion.

8. The method of claim 7, and further comprising prompting verification of the text, based on the marking.

9. The method of claim 1, and further comprising identifying the caller, based on the speech recognition.

10. The method of claim 9, and further comprising selecting fields to be filled, based on the identification of the caller.

11. The method of claim 1, wherein speech recognized from the speech recognition includes at least one of a name, date of birth, and a telephone number.

12. The method of claim 1, wherein the agent includes a human agent.

13. The method of claim 1, wherein the agent includes a software agent.

14. The method of claim 1, wherein the agent includes a call center agent.

15. The method of claim 1, further comprising filling in fields of a database with text generated as a result of the speech recognition.

16. The method of claim 1, further comprising directing a flow of the utterances between the caller and the agent.

17. The method of claim 1, further comprising prompting an action of the agent.

18. The method of claim 1, wherein the context includes grammar.

19. The method of claim 1, wherein the context includes definitions.

22. The method of claim 1, wherein the context is used to recognize the meaning of the portion of the utterances for distinguishing between a set of keywords included in the text having multiple different meanings depending on a context in which the set of keywords is used.

20. A non-transitory computer readable medium embodied on computer program product, comprising computer code when executed by at least one processor to:

sending, via an interface, from an agent to a caller, a request for authentication information for the caller;

receiving, in response to the request for authentication information, a plurality of utterances from the caller, the plurality of utterances including a plurality of spoken words;

performing, by the at least one processor, speech recognition on the plurality of spoken words to generate text representing the utterances;

identifying, by the at least one processor, a context of the utterances, using the text, wherein the context including a particular pattern of the utterances and a particular grammar used in the utterances;

determining, by the at least one processor, a problem and location information associated with the caller using the identified context;

in response to determining that the caller has not provided expected authentication information in the identified context, determining, by at least one processor, whether a solution to the problem is available based on the location information;

when it is determined that the solution to the problem is available, providing, by the at least one processor, the solution to the caller, via the interface, without authenticating the caller;

when it is determined that the solution to the problem is not available, requesting, via the interface, additional authentication information from the caller; and

verifying, by the at least processor, text entered by the agent matches the recognized spoken words; and

in response to detecting a mismatch, alerting the agent, via agent interface, that the mismatch has been detected.

21. A system, comprising:

an interface, to send a request for authentication information from an agent to a caller; and

receive, in response to the request for authentication information, a plurality of utterances from the caller, wherein the plurality of utterances including a plurality of spoken words;

at least one processor, to perform a speech recognition on the plurality of spoken words to generate text representing the utterances;

the at least one processor, to identify a context of the utterances, using the text, wherein the context including a particular pattern of the utterances and a particular grammar used in the utterances;

the at least one processor, to determine a problem and location information associated with the caller using the identified text;

the at least one processor, in response to determining that the caller has not provided expected authentication information, determining whether a solution to the problem is available based on the location information;

when it is determined that the solution to the problem is available, provide, by at least the one processor, the solution to the caller, via the interface, without authenticating the caller; and

when it is determined that the solution to the problem is not available, request, via the interface, additional information from the caller;

verifying, by the at least processor, text entered by the agent matches the recognized spoken words; and

in response to detecting a mismatch, alerting the agent, via agent interface, that the mismatch has been detected.