IP Library › Granted Patent US 10,691,897

Granted Patent B1

US 10,691,897 · App. 16/555,539 · Granted Jun 23, 2020

Artificial intelligence based virtual agent trainer

Inventors: Vidya Rajagopal (Bangalore, IN); Kokila Manickam (Salem, IN); Marin Grace Mercylawrence (Bangalore, IN); Gaurav Mengi (Bangalore, IN)

Assignee: ACCENTURE GLOBAL SOLUTIONS LIMITED

G06F40/35 G06F40/247 G06K9/6257 G06K9/6264 G10L21/18 G10L25/30

View Patent ↗

Loading inventors, assignments & file history…

Monitor This Case

Get email alerts when status or documents change.

Order Certified Copies

Most orders are placed with the USPTO same day — all within 24 business hours.

Order via The Patent Place →

Pre-filled with this patent's details

Quick Facts

Patent No.: US 10,691,897
App. No.: 16/555,539
Granted: Jun 23, 2020
Kind: B1

External Links

USPTO Patent Center↗ Full Patent PDF↗

Abstract

The present disclosure relates to a system, a method, and a product for an artificial intelligence based virtual agent trainer. The system includes a processor in communication with a memory storing instructions. When the processor executes the instructions, the instructions are configured to cause the processor to obtain input data and generate a preliminary set of utterances based on the input data, process the preliminary set of utterances to generate a set of utterance training data, generate a set of conversations based on the set of utterance training data, simulate the set of conversations on a virtual agent to obtain a conversation result, verify an intent and a response based on the conversation result, verify a use case flow and flow hops based on the conversation result, and generate recommendation information and maturity report based on verification results.

Claims (101)

1. A system for computer-based virtual agent trainer, the system comprising:

a memory storing instructions; and

a processor in communication with the memory, wherein, when the processor executes the instructions, the instructions are configured to cause the processor to:

obtain input data comprising a plurality of base utterances,

generate a preliminary set of utterances based on the input data by:

obtaining a plurality of industry keywords based on the input data,

generating a plurality of synonyms of the plurality of industry keywords, and

generating the preliminary set of utterances based on the input data and the plurality of synonyms according to at least one combination of the generated plurality of synonyms of the plurality of industry keywords,

process the preliminary set of utterances to generate a set of utterance training data,

generate a set of conversations based on the set of utterance training data,

simulate the set of conversations on a virtual agent to obtain a conversation result,

verify an intent and a response based on the conversation result,

verify a use case flow and flow hops based on the conversation result, and

generate recommendation information and a maturity report based on verification results.

2. The system of claim 1 , wherein:

the set of utterance training data comprises a set of utterance training data in a text-format and a set of utterance training data in a voice-format; and

when the instructions are configured to cause the processor to process the preliminary set of utterances to generate the set of utterance training data, the instructions are configured to cause the processor to:

process the preliminary set of utterances to generate the set of utterance training data in the text-format, and

generate the set of utterance training data in the voice-format based on the set of utterance training data in the text-format and voice files according a set of voice parameters.

3. The system of claim 2 , wherein:

the voice files comprises at least one of machine-generated voice files or human-generated natural voice files; and

the set of voice parameters comprises at least one of:

a voice type,

a voice volume, or

a voice speed.

4. The system of claim 1 , wherein:

label the preliminary set of utterances, and

check validity of the preliminary set of utterances to obtain the set of utterance training data.

5. The system of claim 1 , wherein, the instructions are configured to cause the processor to generate the set of conversations based on the set of utterance training data, the instructions are configured to cause the processor to:

convert a flow chart to a decision tree;

convert the decision tree to a numeric model;

build a sequence-to-sequence (Seq2Seq) data set based on the decision tree and the numeric model; and

generate the set of conversations based on the set of utterance training data, the Seq2Seq model, and the numeric model.

6. A method for computer-based virtual agent training, the method comprising:

obtaining, by a device comprising a memory storing instructions and system circuitry in communication with the memory, input data comprising a plurality of base utterances;

generating, by the device, a preliminary set of utterances based on the input data by:

obtaining a plurality of industry keywords based on the input data,

generating a plurality of synonyms of the plurality of industry keywords, and

processing, by the device, the preliminary set of utterances to generate a set of utterance training data;

generating, by the device, a set of conversations based on the set of utterance training data;

simulating, by the device, the set of conversations on a virtual agent to obtain a conversation result;

verifying, by the device, an intent and a response based on the conversation result;

verifying, by the device, a use case flow and flow hops based on the conversation result; and

generating, by the device, recommendation information and a maturity report based on verification results.

7. The method of claim 6 , wherein:

the set of utterance training data comprises a set of utterance training data in a text-format and a set of utterance training data in a voice-format; and

the processing the preliminary set of utterances to generate the set of utterance training data comprises:

processing, by the device, the preliminary set of utterances to generate the set of utterance training data in the text-format, and

generating, by the device, the set of utterance training data in the voice-format based on the set of utterance training data in the text-format and voice files according a set of voice parameters.

8. The method of claim 7 , wherein:

the voice files comprises at least one of machine-generated voice files or human-generated natural voice files; and

the set of voice parameters comprises at least one of:

a voice type,

a voice volume, or

a voice speed.

9. The method of claim 6 , wherein:

the processing the preliminary set of utterances to generate the set of utterance training data comprises:

labeling, by the device, the preliminary set of utterances, and

checking, by the device, validity of the preliminary set of utterances to obtain the set of utterance training data.

10. The method of claim 6 , wherein the generating the set of conversations based on the set of utterance training data comprises:

converting, by the device, a flow chart to a decision tree;

converting, by the device, the decision tree to a numeric model;

building, by the device, a sequence-to-sequence (Seq2Seq) data set based on the decision tree and the numeric model; and

generating, by the device, the set of conversations based on the set of utterance training data, the Seq2Seq model, and the numeric model.

11. A product for computer-based virtual agent training, the product comprising:

machine-readable media other than a transitory signal;

instructions stored on the machine-readable media;

a processor in communication with the machine-readable media; and

wherein when the processor executes the instructions, the processor is configured to: