Bot networks

Patent metadata

Field	Value
Publication number	US-10997258-B2
Application number	US-201815908596-A
Country	US
Kind code	B2
Filing date	Feb 28, 2018
Priority date	Feb 28, 2018
Publication date	May 4, 2021
Grant date	May 4, 2021

Abstract

A bot network system may be provided. A system may include a plurality of bot agents, wherein at least one bot agent of the plurality of bot agents is configured to receive a request from a user in natural language. The system may further include a plurality of digital resources including one or more of a software program, a service, a web service and a dataset. Each digital resource of the plurality of digital resources may be configured to communicate with a dedicated bot agent of the plurality of bot agents. Also, each bot agent may be configured to interact with its associated digital resource via an application programming interface (API) of the associated digital resource and translate between the natural language and a language of the associated digital resource.

First claim

What is claimed is:

1. A system, comprising:
a plurality of bot agents, at least one bot agent of the plurality of bot agents configured to receive a user request from a user device in natural language and a plurality of candidate bot agents of the plurality of bot agents configured to obtains responses to the user request; and
a plurality of digital resources including one or more of: a software program, a service, a web service, and a dataset, each digital resource of the plurality of digital resources configured to communicate with a dedicated bot agent of the plurality of candidate bot agents, each candidate bot agent including a trained machine learning model associated with the candidate bot agent and each candidate bot agent configured to:
obtain previous user requests from one or more of the plurality of bot agents, the trained machine learning model associated with the candidate bot agent being-trained using the previous user requests and the one or more of the plurality of bot agents from which the previous user requests are obtained not being included in the plurality of candidate bot agents configured to obtains responses to the user request;
apply the user request to the trained machine learning model associated with the candidate bot agent to obtain a confidence score generated by the trained machine learning model regarding a match between the user request and the one of the plurality of digital resources corresponding to the candidate bot agent;
provide the user request to the one of the plurality of digital resources corresponding to the candidate bot agent;
obtain a response to the user request from the one of the plurality of digital resources corresponding to the candidate bot agent; and
provide the response and the confidence score to the at least one bot agent,

wherein the at least one bot agent is further configured to convey one of the responses from the plurality of candidate bot agents to the user device based on the confidence scores.

2. The system of claim 1 , wherein each of the plurality of candidate bot agents are configured to translate between the natural language of the user request and a language of the corresponding one of the plurality of digital resources.

3. The system of claim 1 , wherein the plurality of bot agents comprise a master bot agent for determining the plurality of candidate bot agents of the plurality of bot agents for handling the user request.

4. The system of claim 1 , wherein at least one bot agent is configured to translate the one of the responses to the natural language of the user request.

5. The system of claim 1 , wherein each bot agent of the plurality of bot agents is configured to exchange its state information with at least one other bot agent of the plurality of bot agents.

6. The system of claim 1 , wherein at least one bot agent of the plurality of bot agents is configured to compare the user request to an API associated with each bot agent to select the plurality of candidate bot agents.

7. A method, comprising:
determining a plurality of candidate bot agents of a plurality of bot agents for responding to a user request from a user device in natural language, each of the plurality of candidate bot agents associated with a different one of a plurality of digital resources and each of the plurality of candidate bot agents including a trained machine learning model, each of the trained machine learning models configured to determine matching between requests and one of the plurality of digital resources;
obtaining, at one or more of the plurality of candidate bot agents, previous user requests from one or more of the plurality of bot agents, the trained machine learning models associated with the one or more of the plurality of candidate bot agents being trained using the previous user requests and the one or more of the plurality of bot agents from which the previous user requests are obtained not being included in the plurality of candidate bot agents for responding to the user request;
conveying, from each of the plurality of candidate bot agents to the corresponding one of the plurality of digital resources, the user request, each of the plurality of digital resources including one or more of: a software program, a service, a web service, and a dataset;
applying, by each of the plurality of candidate bot agents, the user request to their corresponding trained machine learning model configured to determine matching between requests and the one of the plurality of digital resources corresponding to each of the plurality of candidate bot agents;
receiving a plurality of responses from the plurality of digital resources;
selecting one of the plurality of responses as a selected response based on confidences output by the trained machine learning models; and
translating the selected response to the natural language; and
conveying the translated response to the user device.

8. The method of claim 7 , further comprising receiving the user request from the user device in the natural language at a bot agent of the plurality of bot agents.

9. The method of claim 7 , wherein determining the plurality of candidate bot agents comprises determining the plurality of candidate bot agents via a master bot agent.

10. The method of claim 7 , further comprising routing conversation content from a bot agent of the plurality of bot agents to the plurality of candidate bot agents.

11. The method of claim 7 , further comprising exchanging bot agent state information between two or more bot agents of the plurality of bot agents.

12. The method of claim 7 , wherein determining the plurality of candidate bot agents comprises comparing the user request to an API associated with each of the plurality of bot agents.

13. One or more non-transitory computer-readable media that include instructions that, when executed by one or more processors, are configured to cause the one or more processors to perform operations, the operations comprising:
determining a plurality of candidate bot agents of a plurality of bot agents for responding to a user request from a user device in natural language, each of the plurality of candidate bot agents associated with a different one of a plurality of digital resources and each of the plurality of candidate bot agents including a trained machine learning model configured to determine matching between requests and one of the plurality of digital resources;
obtaining, at one or more of the plurality of candidate bot agents, previous user requests from one or more of the plurality of bot agents, the trained machine learning models associated with the one or more of the plurality of candidate bot agents being trained using the previous user requests and the one or more of the plurality of bot agents from which the previous user requests are obtained not being included in the plurality of candidate bot agents for responding to the user request;
conveying, from each of the plurality of candidate bot agents to the corresponding one of the plurality of digital resources, the user request, each of the plurality of digital resources including one or more of: a software program, a service, a web service, and a dataset;
applying, by each of the plurality of candidate bot agents, the user request to their corresponding trained machine learning model configured to determine matching between requests and the one of the plurality of digital resources corresponding to each of the plurality of candidate bot agents;
receiving a plurality of responses from the plurality of digital resources;
selecting one of the plurality of responses as a selected response based on confidences output by the t

Assignees

• Fujitsu Ltd

Inventors

• Chen Wei-Peng
• Liu Lei
• Bahrami Mehdi
• Parikh Samir
• Park Junhee

Classifications

• G06F16/3329 Primary

  Natural language query formulation · CPC title

• G06F9/547

  Remote procedure calls [RPC]; Web services · CPC title

• G06F16/9535 Primary

  Search customisation based on user profiles and personalisation · CPC title