Systems and methods for generating responses for an intelligent virtual

Patent metadata

Field	Value
Publication number	US-11960847-B2
Application number	US-202016835975-A
Country	US
Kind code	B2
Filing date	Mar 31, 2020
Priority date	Apr 4, 2019
Publication date	Apr 16, 2024
Grant date	Apr 16, 2024

Abstract

Systems and methods are described to automatically generate candidate questions and responses to speed the process of response creation and editing for commercial IVAs and chatbots. Rather than create the questions and responses from scratch for a new IVA, the system uses existing questions and responses from a previous or related IVA to train a model that can generate proposed responses to provided questions. The model, or a different model, can further be trained to generate responses using data taken from company or entity-specific data sources such as websites and knowledge bases. After a set of questions and responses have been generated for an IVA they may be reviewed by one or more human reviewers to ensure they are of a suitable quality. Where no previous or related IVA exists to provide example responses, the model may be trained solely using the company or entity-specific data.

First claim

What is claimed:

1. A method for generating responses for use by an intelligent virtual assistant comprising:
receiving a first plurality of responses used by a first intelligent virtual assistant for a first entity by a computing device;
receiving data associated with a second entity by the computing device, wherein the received data is from one or more internal knowledge bases associated with the second entity;
training a machine learning question model using at least some of the received data by the computing device;
generating a plurality of questions by the computing device using the machine learning question model, wherein each question of the plurality of questions simulates a question from a customer of the second entity to the first intelligent virtual assistant;
assigning an intent to each question of the plurality of questions by the computing device;
training a first machine learning response model using at least some of the received data by the computing device and without using the first plurality of responses;
training a second machine learning response model to generate responses for use by the second intelligent virtual assistant using at least some of the first plurality of responses used by the first intelligent virtual assistant and without using the received data by the computing device;
generating a plurality of responses using the first machine learning response model, the second machine learning response model, each question from the plurality of questions, and the assigned intents by the computing device; and
providing the questions of the plurality of questions, the assigned intents, and the generated plurality of responses, for use by a second intelligent virtual assistant for the second entity by the computing device.

2. The method of claim 1 , further comprising providing the plurality of questions to a human reviewer.

3. The method of claim 1 , wherein assigning an intent to each question comprises generating a mapping of questions to intents.

4. The method of claim 3 , further comprising providing the mapping to a human reviewer.

5. The method of claim 1 , further comprising providing the generated responses to a human reviewer.

6. The method of claim 1 , further comprising transforming the received data into a format used by the machine learning question model or the first machine learning response model.

7. The method of claim 1 , wherein assigning an intent to a question comprises providing the question to the first intelligent virtual assistant and receiving the intent from the first intelligent virtual assistant.

8. The method of claim 1 , further comprising scraping the received data from one or more knowledge bases associated with the second entity.

9. The method of claim 1 , wherein the first plurality of responses is received from one or more of the first intelligent virtual assistant, one or more call transcripts, or the one or more internal knowledge bases.

10. A system for generating responses for use by an intelligent virtual assistant comprising:
at least one processor; and
a non-transitory computer readable medium comprising instructions that, when executed by the at least one processor, cause the system to:
receive a first plurality of responses used by a first intelligent virtual assistant for a first entity;
receive data associated with a second entity, wherein the received data is from one or more internal knowledge bases associated with the second entity;
train a machine learning question model using at least some of the received data;
generate a plurality of questions using the machine learning question model, wherein each question of the plurality of questions simulates a question from a customer of the second entity to the first intelligent virtual assistant;
assign an intent to each question of the plurality of questions;
train a first machine learning response model using at least some of the received data;
train a second machine learning response model to generate responses for use by a second intelligent virtual assistant using at least some of the first plurality of responses used by the first intelligent virtual assistant and without using the received data;
generate a plurality of responses using the first machine learning response model, the second machine learning response model, each question from the plurality of questions, and the assigned intents by the computing device; and
provide the questions of the plurality of questions, the assigned intents, and the generated plurality of responses, or use by the second intelligent virtual assistant for the entity.

11. The system of claim 10 , further comprising instructions that, when executed by the at least one processor, cause the system to provide the plurality of questions to a human reviewer.

12. The system of claim 10 , wherein assigning an intent to each question comprises generating a mapping of questions to intents.

13. The system of claim 12 , further comprising instructions that, when executed by the at least one processor, cause the system to provide the mapping to a human reviewer.

14. The system of claim 10 , further comprising instructions that, when executed by the at least one processor, cause the system to provide the generated plurality of responses to a human reviewer.

15. The system of claim 10 , further comprising instructions that, when executed by the at least one processor, cause the system to transform the received data into a format used by the machine learning question model or the first machine learning response model.

16. A non-transitory computer readable medium comprising instructions that, when executed by at least one processor, cause a system to:
receive a first plurality of responses used by a first intelligent virtual assistant for a first entity;
receive data associated with a second entity, wherein the received data is from one or more internal knowledge bases associated with the second entity;
train a machine learning question model using at least some of the received data;
generate a plurality of questions using the machine learning question model, wherein each question of the plurality of questions simulates a question from a customer of the second entity to the first intelligent virtual assistant;
assign an intent to each question of the plurality of questions;
train a first machine learning response model using at least some of the received data;
train a second machine learning response model to generate responses for use by a second intelligent virtual assistant using at least some of the first plurality of responses used by the first intelligent virtual assistant and without using the received data;
generate a plurality of responses using the first machine learning response model, the second machine learning response model, each question from the plurality of questions, and the assigned intents by the computing device; and
provide the questions of the plurality of questions, the assigned intents, and the generated plurality of responses, for use by the second intelligent virtual assistant for the entity.

17. The computer-readable medium of claim 16 , further comprising instructions that, when executed by the at least one processor, cause the system to provide the plurality of questions to a human reviewer.

18. The computer-readable medium of claim 16 , wherein assigning an intent to each question comprises generating a mapping of questions to intents.

19. The computer-readable medium of claim 18 , further comprising instructions that, when executed by t

Assignees

• Verint Americas Inc

Inventors

• Beaver Ian

Classifications

• G06F40/35 Primary

  Discourse or dialogue representation · CPC title

• G06F16/90332

  Natural language query formulation or dialogue systems · CPC title

• G06F18/2178

  based on feedback of a supervisor · CPC title

• G06N20/00

  Machine learning · CPC title

• G06N5/04

  Inference or reasoning models · CPC title