Artificial intelligence oriented permission management method and system

Abstract

An AI-driven access control method and system are disclosed. In this method, a workflow is first established based on a task, and multiple work nodes are created, along with an intelligent robot assigned to perform a stage task at each work node. Then, during workflow execution, an AI-driven access control model is used to verify the permissions of the multiple intelligent robots corresponding to the multiple work nodes before sequentially executing the stage tasks of each work node. Next, after the intelligent robots corresponding to each work node complete their stage tasks, the AI-driven access control model verifies the permissions of the intelligent robot corresponding to the next work node before continuing to execute the stage tasks of the next work node.

Claims

1. An AI-driven access control method, implemented in a computer system, comprising: Establish a workflow based on a task; establish multiple work nodes based on the workflow, and assign an intelligent robot to perform the stage task at each work node; When executing this workflow, an artificial intelligence permission management model is used to confirm the permissions of multiple intelligent robots corresponding to the multiple work nodes before sequentially executing the phase tasks of each work node; wherein, after the intelligent robot corresponding to each work node completes the phase task of that work node, the artificial intelligence permission management model is used to confirm the permissions of the intelligent robot corresponding to the next work node before continuing to execute the phase task of the next work node.

2. The AI-driven access control method as described in claim 1, wherein the AI ​​access control model runs in a robot agent system, the robot agent system receives the task submitted by a client, and accepts the conditions set by the client to execute the task through a user interface, and after analyzing the task, establishes the workflow with multiple work nodes, each work node having an intelligent robot that performs the task in the execution phase.

3. The AI-driven access control method as described in claim 2, wherein the robot agent system runs a search engine, connects to a robot service system, and when receiving the task and establishing the workflow, retrieves the multiple intelligent robots through the search engine based on the conditions of the workflow.

4. The AI-guided access control method as described in request item 3, wherein, The robot service system includes a robot database, in which a machine learning algorithm is used to train multiple intelligent robots by learning from data on multiple domains and multiple job contents, and then deploys them into the robot service system.

5. The AI-driven access control method as described in claim 3, wherein the search engine searches for multiple intelligent robots that meet the conditions of the workflow and can perform stage tasks at multiple work nodes based on one or any combination of the functions, costs, evaluations, and permissions of each intelligent robot.

6. The AI-driven access control method as described in claim 3, wherein, After the intelligent robot at each work node completes its phase task, the AI ​​permission management model determines that the intelligent robot at the next work node does not meet the permission management settings. Then, the search engine searches the robot service system to adjust the intelligent robot.

7. The AI-driven access control method as described in Request 1, wherein when establishing the workflow, a timetable for completing the phase tasks of each work node in the workflow is also established.

8. The AI-oriented access control method as described in any one of claims 1 to 7, wherein the AI ​​access control model based on a large language model is used to perform access control that simulates human social hierarchy for the multiple intelligent robots performing the task, in order to handle access control for authentication, authorization and access control at each stage of the workflow.

9. The AI-guided access control method as described in any one of claims 1 to 7, wherein the method for training the AI ​​access control model includes first preparing training data on human social operation and class relations, and then learning the interpersonal access control relationships in the training data through a deep learning method, wherein the AI ​​access control model is established by learning the class relations between people in the data using a neural network that mimics human beings, and is used to handle access control for authentication, authorization and access control of tasks at each stage of the workflow.

10. A system for implementing an AI-driven access control method, comprising: A processor and a temporary memory, wherein the temporary memory is used to temporarily store information of multiple intelligent robots performing a task; The processor runs an AI-driven access control model to implement the AI-oriented access control method, which includes: establishing a workflow based on the task; establishing multiple work nodes based on the workflow, and assigning an intelligent robot to perform a stage task at each work node; and when executing the workflow, using the AI-driven access control model to confirm the permissions of the multiple intelligent robots corresponding to the multiple work nodes, and then sequentially executing the stage tasks of each work node; wherein, after the intelligent robot corresponding to each work node completes the stage task of that work node, the AI-driven access control model confirms the permissions of the intelligent robot corresponding to the next work node, and then continues to execute the stage task of the next work node.

11. The system as described in claim 10, wherein the artificial intelligence access control model operates in a robot agent system, the robot agent system receives the task submitted by a customer, and accepts the conditions set by the customer to execute the task through a user interface, and after analyzing the task, establishes the workflow with the multiple work nodes, each work node having an intelligent robot that performs the task in the execution phase.

12. The system as described in claim 11, wherein the robot agent system operates a search engine, connects to a robot service system, and, upon receiving the task and establishing the workflow, obtains the plurality of intelligent robots through the search engine based on the conditions of the workflow.

13. The system as described in claim 12, wherein, The robot service system includes a robot database, in which a machine learning algorithm is used to train multiple intelligent robots by learning from data on multiple domains and multiple job contents, and then deploys them into the robot service system.

14. The system as described in claim 12, wherein, After the intelligent robot at each work node completes its phase task, the AI ​​permission management model determines that the intelligent robot at the next work node does not meet the permission management settings. Then, the search engine searches the robot service system to adjust the intelligent robot.

15. The system of any one of claims 10 to 14, wherein the artificial intelligence access control model based on a large language model is used to perform access control that simulates human social hierarchy for the plurality of intelligent robots performing the task, in order to handle access control for authentication, authorization and access control of each stage of the workflow.

16. The system of any one of claims 10 to 14, wherein the method of training the artificial intelligence access control model comprises first preparing training data on human social operation and class relations, and then learning the interpersonal access control relationships in the training data through a deep learning method, wherein the artificial intelligence access control model is established by learning the interpersonal class relations in the data using a neural network that mimics human beings, for handling access control management of authentication, authorization and access control for performing tasks at each stage of the workflow.

Images