Granted Patent B2

US 12,197,860 · App. 17/383,631 · Granted Jan 14, 2025

Inventors: Prakash Ranganathan (Tamil Nadu, IN); Saurabh Tahiliani (Noida, IN)

Assignee: Verizon Patent and Licensing Inc.

G06F40/226G06F40/35G06F40/56G06N20/20

Abstract

Claims (82)

1. A system comprising:

a processor configured to execute instructions to facilitate performance of operations comprising:

identifying a conversation path associated with a revised code segment of a conversational interaction entity, the conversation path having a predetermined intent;

generating a conversational phrase for the conversation path;

employing the conversational interaction entity using the conversation path and the conversational phrase to generate a resultant intent;

determining that the resultant intent generated by the conversational interaction entity using the conversation path does not match the predetermined intent of the conversation path;

generating an issue report for the conversational interaction entity responsive to the determination that the resultant intent does not match the predetermined intent;

comparing a first conversation path to a second conversation path to identify a common path that does not differ between the first conversation path and the second conversation path and a sub-path that differs between the first conversation path and the second conversation path;

generating context information for the sub-path; and

employing the conversational interaction entity using the context information to generate a sub-path resultant intent for the sub-path separate from the common path.

2. The system of claim 1 , wherein:

identifying the conversation path comprises:

employing a first machine learning engine to identify the conversation path.

3. The system of claim 2 , the operations comprising:

training the first machine learning engine using reference code segments, each having an associated reference conversation path; and

employing the first machine learning engine to identify the conversation path based on the revised code segment.

4. The system of claim 1 , wherein:

generating the conversational phrase comprises:

receiving a first seed phrase associated with the conversation path; and

employing a second machine learning engine to generate the conversational phrase using alternative language for the first seed phrase.

5. The system of claim 4 , the operations comprising:

determining a confidence metric associated with the alternative language and the resultant intent; and

responsive to the confidence metric exceeding a threshold, designating the alternative language as a second seed phrase for the conversation path.

6. The system of claim 5 , the operations comprising:

generating an entry in the issue report responsive to the confidence metric associated with the alternative language not exceeding the threshold.

7. The system of claim 1 , wherein:

employing the conversational interaction entity using the conversation path to generate the resultant intent comprises:

calling an application programming interface of the conversational interaction entity using the conversation path and the conversational phrase to generate the resultant intent.

8. The system of claim 1 , the operations comprising:

receiving a response from the conversational interaction entity for the resultant intent;

performing a language validation on the response; and

generating an entry in the issue report responsive to the response failing the language validation.

9. A non-transitory machine-readable medium storing instructions that when executed facilitate performance of operations comprising:

identifying a conversation path associated with a revised code segment of a conversational interaction entity, the conversation path having a predetermined intent;

generating a conversational phrase for the conversation path;

employing the conversational interaction entity using the conversation path and the conversational phrase to generate a resultant intent;

determining that the resultant intent generated by the conversational interaction entity using the conversation path does not match the predetermined intent of the conversation path;

generating an issue report for the conversational interaction entity responsive to the determination that the resultant intent does not match the predetermined intent;

comparing a first conversation path to a second conversation path to identify a common path that does not differ between the first conversation path and the second conversation path and a sub-path that differs between the first conversation path and the second conversation path;

generating context information for the sub-path; and

employing the conversational interaction entity using the context information to generate a sub-path resultant intent for the sub-path separate from the common path.

10. The non-transitory machine-readable medium of claim 9 , wherein:

identifying the conversation path comprises:

employing a first machine learning engine to identify the conversation path.

11. The non-transitory machine-readable medium of claim 10 , the operations comprising:

training the first machine learning engine using reference code segments, each having an associated reference conversation path; and

employing the first machine learning engine to identify the conversation path based on the revised code segment.

12. The non-transitory machine-readable medium of claim 9 , wherein:

generating the conversational phrase comprises:

receiving a first seed phrase associated with the conversation path; and

employing a second machine learning engine to generate the conversational phrase using alternative language for the first seed phrase.

13. The non-transitory machine-readable medium of claim 12 , the operations comprising:

determining a confidence metric associated with the alternative language and the resultant intent; and

responsive to the confidence metric exceeding a threshold, designating the alternative language as a second seed phrase for the conversation path.

14. The non-transitory machine-readable medium of claim 13 , the operations comprising:

generating an entry in the issue report responsive to the confidence metric associated with the alternative language not exceeding the threshold.

15. The non-transitory machine-readable medium of claim 9 , wherein:

employing the conversational interaction entity using the conversation path to generate the resultant intent comprises:

calling an application programming interface of the conversational interaction entity using the conversation path and the conversational phrase to generate the resultant intent.

16. The non-transitory machine-readable medium of claim 9 , the operations comprising:

receiving a response from the conversational interaction entity for the resultant intent;

performing a language validation on the response; and

generating an entry in the issue report responsive to the response failing the language validation.

17. A method comprising:

identifying, by a processor, a conversation path associated with a revised code segment of a conversational interaction entity, the conversation path having a predetermined intent;

generating, by the processor, a conversational phrase for the conversation path;

employing, by the processor, the conversational interaction entity using the conversation path and the conversational phrase to generate a resultant intent;

determining, by the processor, that the resultant intent generated by the conversational interaction entity using the conversation path does not match the predetermined intent of the conversation path;

generating, by the processor, an issue report for the conversational interaction entity responsive to the determination that the resultant intent does not match the predetermined intent;

comparing, by the processor, a first conversation path to a second conversation path to identify a common path that does not differ between the first conversation path and the second conversation path and a sub-path that differs between the first conversation path and the second conversation path;

generating, by the processor, context information for the sub-path; and

employing, by the processor, the conversational interaction entity using the context information to generate a sub-path resultant intent for the sub-path separate from the common path.

18. The method of claim 17 , wherein:

identifying, by the processor, the conversation path comprises:

employing, by the processor, a first machine learning engine to identify the conversation path.

19. The method of claim 18 , comprising:

training, by the processor, the first machine learning engine using reference code segments, each having an associated reference conversation path; and

employing, by the processor, the first machine learning engine to identify the conversation path based on the revised code segment.

20. The method of claim 17 , wherein:

generating, by the processor, the conversational phrase comprises:

receiving, by the processor, a first seed phrase associated with the conversation path; and

employing, by the processor, a second machine learning engine to generate the conversational phrase using alternative language for the first seed phrase.

Assignments (1)

Continuity (1)

References Cited (9)