Governance-based autonomous processing system

Abstract

This invention provides a governed autonomous processing system for handling requests or events from human users or external systems within a governed application environment. The system comprises a cognitive core module, an experiential core module, and a behavioral core module. These three modules are logically distinct and have a forward-directed permission restriction relationship during governance, but a reverse feedback relationship during learning. During governance, the cognitive core module is responsible for request parsing and governance judgment; the experiential core module is responsible for task flow management and conditional clarification; and the behavioral core module executes actions at least when governance permits and the task is established. However, upon completion of an action, the result can be fed back from the behavioral core to the experiential core, then filtered and extracted by the experiential core before being fed back to the cognitive core for information expansion and updating. The disclosed method ensures the governability, security, and auditability of autonomous processing behavior by restricting only results that pass governance to being incorporated into learning or experience processing.

Description

Governance-based autonomous processing system Technical Field

[0001] This work relates to an autonomous processing system, particularly a governed autonomous processing system in governed application scenarios that restricts the execution of behavior through a governance judgment mechanism. Prior Technology

[0002] With the development of information systems, autonomous control technologies, and intelligent applications, existing systems can now process requests or events from human users or external systems through rule engines, automated processes, or intelligent agent mechanisms, and automatically execute corresponding entity or digital behaviors. Such systems often focus on improving automation or real-time response capabilities in their design, and attempt to integrate decision-making logic, task management, and behavior execution functions into a single processing architecture.

[0003] However, when applied in regulated or high-responsibility settings, the aforementioned technologies often have the following limitations: First, the distinction between action execution and governance judgment is not clear, which may lead to the system executing actions without sufficient evaluation or authorization; second, task processes are usually regarded as the direct result of requests, lacking flexible mechanisms for suspension, clarification, or re-evaluation; third, the results of actions are easily incorporated into subsequent learning or experience accumulation, increasing the risk of internalizing erroneous or unauthorized actions.

[0004] Furthermore, existing technologies often fail to clearly define the boundaries of authority at different processing stages, making it difficult to clarify the attribution of responsibility among governance judgments, task coordination, and action execution, which is also detrimental to subsequent auditing or risk control. Therefore, how to establish a technical architecture that can clearly restrict action execution, support conditional clarification, and ensure that the learning process is subject to governance constraints without sacrificing autonomous processing capabilities still requires further improvement. Summary of the Invention

[0005] This invention discloses a governance-based autonomous processing system for handling requests or events from human users or external systems within a governed application environment, and executing corresponding entity or digital actions when governance conditions are met. The system comprises at least a cognitive core module, an experiential core module, and a behavioral core module, which are logically distinct from each other and have a forward-directed permission restriction relationship and a reverse feedback relationship. The execution of any action requires confirmation of the aforementioned governance judgment result and task flow as a prerequisite; however, this feedback does not affect the unidirectional permission relationship of the governance judgment.

[0006] The cognitive core module is configured to receive and parse the request or event to form a structured request description, identify the source role, the state of the domain, and the corresponding regulatory conditions of the request or event, and then, based on at least one set of governance regulations, role permission information, domain restriction information, and system capability boundaries, perform governance judgments on the structured request description and output one of the following governance judgment results: allow, deny, or conditional. This cognitive core module does not perform any entity or digital actions, nor does it directly invoke external tools or capability modules.

[0007] This core experience module is configured to establish, maintain, update, or terminate a task process based on the aforementioned governance judgment results. The establishment of this task process is not a necessary outcome of the request or event. When the governance judgment result is conditional, the core experience module further initiates an interactive process of requesting supplementary information or verification to obtain the supplementary information required for re-judging the governance and manages the status changes of the task process during execution. Only when the task process passes the governance conditions does the core experience module decide whether to include its execution or processing results in subsequent experience compilation or learning data sources. This core experience module does not have governance judgment authority and does not directly execute any actions. This behavior core module is configured to execute the corresponding entity or digital behavior only when it receives a request from the experience core module and the corresponding task flow has been approved by the governance mechanism. After the behavior is executed, it sends the behavior result and diagnostic information back to the experience core module. This behavior core module does not have the authority to initiate governance decisions, modify governance decision results, or bypass the governance decision to execute behaviors directly. Simple Explanation of the Diagram

[0008] Figure 1 is a schematic diagram of the information flow of the governance-based autonomous processing system during the task execution phase in one embodiment of this invention. Figure 2 is a schematic diagram of the information flow of the governance-based autonomous processing system during the learning or experience feedback stage in one embodiment of this invention. Figure 3 is a schematic diagram of the system composition and module configuration of the governance-based autonomous processing system in one embodiment of this invention. Implementation

[0009] The embodiments of this invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this invention, and should not be construed as limiting this invention.

[0010] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential," indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing this invention and simplifying the description. They do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on this invention. Furthermore, features specified as "first" or "second" may explicitly or implicitly include one or more of those features. In the description of this invention, unless otherwise stated, "plural" means two or more.

[0011] In the description of this invention, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," "joining," and "assembly" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to connections within two components. Those skilled in the art can understand the specific meanings of the above terms in this invention based on the specific circumstances.

[0012] Referring to Figures 1 and 2, this invention proposes a governance-based autonomous processing system 100, which is used in a governed application environment to process requests or events from human users or external systems, and to perform corresponding physical or digital actions when governance conditions are met. In this embodiment, the governance-based autonomous processing system 100 includes at least a cognitive core module 10, an experience core module 20, and a behavior core module 30, wherein each module interacts according to different information transmission relationships during the task execution phase and the learning or experience feedback phase.

[0013] Referring to Figure 3, the governance-based autonomous processing system 100 can be implemented in a hardware device, which includes the cognitive core module 10, the experience core module 20, and the behavior core module 30. These modules are configured in the same device and are logically distinct from each other to jointly constitute the governance-based autonomous processing system 100.

[0014] The cognitive core module 10 is configured to receive the request or event and parse it to form a structured request description, further identifying the source role, the field status, and the corresponding regulatory conditions of the request or event. Furthermore, the cognitive core module 10 performs governance judgment on the structured request description based on at least one set of governance regulatory data, role permission information, field restriction information, and system capability boundaries, and outputs a governance judgment result, which includes at least one of three states: allow, deny, or conditional. In this embodiment, the cognitive core module 10 does not perform any entity or digital actions, nor does it directly call external tools or capability modules; it is only responsible for governance judgment and request understanding.

[0015] In this embodiment, the governance-related data used within the cognitive core module 10 is managed hierarchically according to its nature or governance level, including at least an immutable core governance data level and a non-core data level that can be updated based on learning or experience feedback. The immutable core governance data level stores data related to governance norms, permission boundaries, or basic system governance principles, and its content is not modified by the task flow processing results or learning organization of the experience core module 20. The updatable non-core data level can be supplemented or adjusted based on the organization results fed back by the experience core module 20 when governance conditions are met.

[0016] The experience core module 20 is configured to establish, maintain, update, or terminate a task process based on the aforementioned governance judgment result, wherein the establishment of the task process is not a necessary result of the request or event. When the governance judgment result is conditional, the experience core module 20 initiates a supplementary information request or verification interaction process to obtain the supplementary information required for re-performing the governance judgment. The experience core module 20 also manages the state changes of the task process during execution, and its state includes at least one or a combination of establishment, termination, or completion. Only when the task process passes the governance conditions does the experience core module 20 decide whether to include its execution result or processing result in subsequent experience compilation or learning data sources. In this embodiment, the experience core module 20 does not have governance judgment authority, nor does it directly execute any entity or digital behavior.

[0017] The behavior core module 30 is configured to execute the corresponding entity or digital behavior only when it receives a request from the experience core module 20 and the corresponding task flow has been approved by the governance judgment. After the behavior is executed, it sends the behavior result and diagnostic information back to the experience core module 20. In this embodiment, the behavior core module 30 does not have the authority to initiate governance judgment, modify governance judgment results, or bypass the governance judgment to directly execute the behavior.

[0018] Furthermore, the cognitive core module 10, the experience core module 20, and the behavior core module 30 are logically distinct from each other and have a forward permission restriction relationship. The execution of any behavior must be based on the aforementioned governance judgment result and task process confirmation as a prerequisite, so as to ensure that when the system executes behavior in the governed application field, it can maintain the technical characteristic that governance judgment precedes behavior execution.

[0019] In this embodiment, the governance-based autonomous processing system 100 has clearly distinguished information transmission directions at different processing stages. Specifically, during the task execution stage, requests or events are sequentially processed by the cognitive core module 10 and then transmitted to the experience core module 20 to establish or confirm the corresponding task flow. When the task flow passes the governance conditions, the behavior core module 30 executes the corresponding entity or digital behavior, forming a forward processing flow from the cognitive core module 10 through the experience core module 20 to the behavior core module 30.

[0020] Conversely, during the learning or experience feedback phase, the behavioral results or diagnostic information generated by the behavioral core module 30 after the behavior is executed are first sent back to the experience core module 20 for sorting, screening or extraction. When the governance conditions are met, the information is then fed back to the cognitive core module 10 for subsequent governance judgment or cognitive data updates, forming a reverse feedback process from the behavioral core module 30 through the experience core module 20 to the cognitive core module 10.

[0021] In this embodiment, the three states of the aforementioned governance judgment result correspond to different processing methods. When the governance judgment result is in an allowed state, the experience core module 20 allows the task flow to enter the subsequent behavior execution stage, and the behavior core module 30 executes the corresponding entity or digital behavior. When the governance judgment result is in a rejected state, the experience core module 20 terminates the task flow, and the system does not execute any entity or digital behavior. When the governance judgment result is in a conditional state, the experience core module 20 suspends the task flow and waits for supplementary information or verification results to be obtained before re-performing the governance judgment to determine the subsequent processing method.

[0022] In this embodiment, when the aforementioned governance judgment result is conditional, the experience core module 20 is configured to send at least one supplementary information request or verification request to the human user or external system to obtain the supplementary information or verification result required for re-performing the governance judgment. After obtaining the supplementary information or verification result, the experience core module 20 incorporates it into the subsequent request parsing or governance judgment process so that the cognitive core module 10 can perform the governance judgment again.

[0023] In this embodiment, the behavior core module 30 is configured not to execute any entity or digital behavior before obtaining a governance judgment result. That is, before the governance judgment result confirms a permitted state, the behavior core module 30 is restricted from executing any behavior to prevent ungoverned behavior from being executed and to ensure that the system's behavior execution is always constrained by the governance judgment.

[0024] In this embodiment, only the execution results or processing results of task processes that pass the governance judgment are included in the experience compilation or learning data source. Conversely, the execution results or processing results of task processes that fail the governance judgment, are rejected, or are terminated are not included in subsequent learning or experience compilation processes, in order to avoid errors, overreach, or inappropriate behavior affecting the system's subsequent processing behavior.

[0025] In this embodiment, the cognitive core module 10, the experience core module 20, or the behavior core module 30 generates corresponding structured records at at least one key node of governance judgment, task process change, or behavior execution. These structured records can be used for subsequent traceability, auditing, or accountability clarification to support the auditability of the system's operation in the governed application environment.

[0026] In this embodiment, the invention further discloses a governance-based autonomous processing method for handling requests or events from human users or external systems in a governed application environment. This method includes at least the following steps: request reception and parsing, role and environment identification, governance judgment, task flow determination, conditional clarification, behavior execution permission determination, behavior execution and result feedback, and experience organization and learning constraints.

[0027] In the request reception and parsing step, the system receives the request or event and parses it to form a structured request description. Following this, in the role and field identification step, the system identifies the source role of the request or event, its field status, and the corresponding specification conditions.

[0028] In the governance judgment step, the system performs governance judgment on the aforementioned structured request description based on at least one governance specification document, role and permission information, field restriction information, and system capability boundaries, and outputs a governance judgment result, wherein the governance judgment result includes at least one of three states: allow, deny, or conditional.

[0029] In the task flow decision-making step, the system determines whether to establish, maintain, update, or terminate a task flow based on the aforementioned governance judgment results. The establishment of the task flow is not a necessary outcome of the request or event. When the governance judgment result is conditional, the system enters the conditional clarification step, initiating an interactive process of requesting supplementary information or verification to obtain the supplementary information required for re-performing the governance judgment.

[0030] In the action execution permission determination step, the system only allows the execution of the corresponding entity or digital action if the governance judgment result is permitted and the corresponding task flow is confirmed to be valid. In the action execution and result feedback step, the system executes the corresponding entity or digital action and, after the action execution is completed, feeds back the action result and diagnostic information for subsequent processing.

[0031] In the experience organization and learning restriction steps, the system only decides whether to include the execution result or processing result in the subsequent experience organization or learning data source when the task process passes the governance conditions.

[0032] In this embodiment, in the aforementioned conditional clarification step, after the system obtains the supplementary information or verification result, the governance-based autonomous processing method further includes re-executing the request reception and parsing step, the role and field identification step, and the governance judgment step, so as to perform governance judgment again based on the updated structured request description, thereby determining the subsequent task flow processing method.

[0033] In this embodiment, when the governance judgment result output in the aforementioned governance judgment step is rejection, or when the task process is terminated in the task process decision step, the governance-based autonomous processing method does not execute the aforementioned behavior execution permission judgment step and behavior execution and result feedback step, so as to prevent entities or digital behaviors that have not passed governance from being executed.

[0034] In this embodiment, in the aforementioned experience organization and learning restriction steps, the system will only include the corresponding execution result or processing result in the subsequent experience organization or learning data source if the governance judgment result is allowed and the task process is completed; and if the governance judgment result is rejected or conditional, the corresponding processing result will not be used as learning data.

[0035] Furthermore, in another embodiment, the governance-based autonomous processing system and method of this invention can be applied to request processing scenarios within an enterprise's internal information system, such as requests from employees for data access, system operation, or cross-departmental collaboration. In this application scenario, the system can make governance judgments based on user roles, organizational levels, and internal regulations, and only allow corresponding system operations or data processing behaviors when the governance conditions are met.

[0036] Furthermore, in another embodiment, this invention can be applied to smart devices or IoT management environments to process events or instructions from sensing devices, control devices, or external management platforms. In this scenario, the system can first perform a governance judgment to confirm whether the event or instruction falls within the permitted operating range and prevent ungoverned device control or automated behavior from being executed.

[0037] Furthermore, in another embodiment, this invention can be applied to interactive request processing scenarios of digital assistants, virtual service agents, or self-service platforms. In this application scenario, the system can identify the user's role and determine the context of the request, and if the determination result is conditional, initiate a supplementary information request or verification process to avoid the system directly executing actions when information is insufficient or permissions are unclear.

[0038] Furthermore, in another embodiment, this invention can be applied to financial, healthcare, public service, or other highly regulated application scenarios to handle requests or events involving data access, transaction execution, or service initiation. In such application scenarios, the system, through a design that prioritizes governance judgment over action execution, ensures that only actions meeting regulatory conditions are executed, and that the subsequent processing and learning of action results are subject to governance constraints.

[0039] Furthermore, in another embodiment, this invention can be applied to systems with autonomous learning or experience accumulation mechanisms, in which only the processing results that pass the governance judgment are included in subsequent experience organization or learning data sources, while requests, events, or behavioral results that do not pass the governance judgment are not used as learning basis, so as to avoid inappropriate behavior being internalized into system capabilities.

[0040] Furthermore, in another embodiment, the governance-based autonomous processing method of this invention can be applied to service robot systems in semi-public spaces, such as counter service robots in residential communities, buildings, or parks, to handle requests or events from different roles.

[0041] In this application scenario, the method involves receiving requests or events from residents, visitors, or administrators, and parsing the requests while identifying roles and locations. A governance judgment step is then executed to determine whether the request complies with community management guidelines, role permissions, and location restrictions. If the governance judgment result is conditional, the method further executes a conditional clarification step to request supplementary information or confirm authorization status from the user. Only if the governance judgment result is "allowed" and the task flow is confirmed, is the corresponding service action permitted; if the governance judgment result is "rejected," no action is taken. Only the results of successful governance processes can be incorporated into subsequent experience compilation or learning materials.

[0042] Furthermore, in another embodiment, the governance-based autonomous processing method of this invention can be applied to event-oriented task processing systems, such as application scenarios for package collection, package anomaly, or security incident reporting, where the event can be triggered by humans, sensing devices, or external systems.

[0043] In this application scenario, the method includes receiving an event and performing event parsing and structuring to identify the event type and related role information, and executing a governance judgment step to determine whether to allow the establishment of a corresponding task flow. When the governance judgment result is conditional, the method further executes a conditional clarification step to confirm the nature of the event or the authorization status; the corresponding task flow is established and executed only if the governance judgment result is allowable; if the governance judgment result is denyable, no action is taken. This method also restricts the inclusion of only the results of event processing that have passed governance into subsequent experience compilation or learning materials.

[0044] Furthermore, in another embodiment, the governance-based autonomous processing method of this invention can be applied to retail or commercial transaction settings, such as self-checkout, unmanned stores, or smart checkout systems, to process transaction requests involving regulatory restrictions or liability risks.

[0045] In this application scenario, the method includes receiving transaction requests and parsing transaction content, identifying user roles, product attributes, and site regulations, and performing governance judgment steps to determine whether the transaction is permitted. When the governance judgment result is conditional, the method further performs a conditional clarification step to request the user to provide identity verification or other supplementary information; the corresponding transaction behavior is only allowed when the governance judgment result is updated to permitted and the task flow is confirmed to be established; if the governance judgment result is rejected, no transaction behavior is executed. This method also restricts the inclusion of only transactions that pass governance into subsequent experience compilation or learning data sources.

[0046] Furthermore, in another embodiment, the governance-based autonomous processing method of this invention can be applied to conflict resolution scenarios in the same field where multiple roles make different requests for the same event.

[0047] In this application scenario, the method involves simultaneously receiving, parsing, and associating multiple requests, identifying the role, source, and context of each request, and then executing governance judgment steps to determine whether each request is allowed, rejected, or requires a conditional clarification process. This method allows for different governance judgment results between different requests, and only requests with a allowed governance judgment result will have their corresponding task processes established and executed; other requests will not have tasks established or actions executed. Only the results of governance processing can be incorporated into subsequent experience compilation or learning materials.

[0048] Furthermore, in another embodiment, the governance-based autonomous processing method of this invention can be applied to digital assistants, governed agent systems, or tool scheduling systems to handle requests involving external tools, models, or services.

[0049] In this application scenario, the method involves receiving user requests, parsing and performing governance judgments, and only allowing the execution of corresponding tool calls or service requests when the governance judgment result is permitted and the task flow is confirmed to be valid. When the governance judgment result is conditional or denial, the method prohibits the execution of any tools or capabilities and prevents the results of failing governance from being included in subsequent experience compilation or learning materials.

[0050] While this work discloses the above embodiments, it is not intended to limit this work. Any modifications and simple substitutions made by those skilled in the art without departing from the spirit and scope of this work are covered by the claims of this work and their equivalents.

[0051] 100: Governance-based Autonomous Processing System

[0052] 10: Cognitive Core Module

[0053] 20: Experience Core Module

[0054] 30: Core Behavioral Module

Claims

1. A governance-based autonomous processing system, used in a governed application domain to process requests or events from human users or external systems, and to execute corresponding actions when governance conditions are met, characterized in that the system at least includes: a cognitive core module (10), the cognitive core module (10) being configured to: receive the request or event and parse it to form a structured request description; identify the source role, domain status, and corresponding regulatory conditions of the request or event; perform governance judgment on the structured request description based on at least one set of governance regulatory data, role permission information, domain restriction information, and system capability boundaries, and output a governance judgment result, wherein the governance judgment result includes at least one of the following three states: allow, deny, or conditional; wherein, The cognitive core module (10) does not perform any entity or digital behavior, nor does it directly call external tools or capability modules; An experience core module (20) is connected to the cognitive core module (10), and the experience core module (20) is configured to: establish, maintain, update or terminate a task process based on the governance judgment result, wherein the establishment of the task process is not a necessary result of the request or event; when the governance judgment result is conditional, initiate an interactive process of supplementary information request or verification to obtain the supplementary information required for re-performing the governance judgment; manage the status changes of the task process during execution, including at least one or a combination of establishment, termination or completion; only when the task process passes the governance conditions, decide whether to include its execution result or processing result in subsequent experience compilation or learning data sources; wherein the experience core module (20) does not have governance judgment authority and does not directly perform any behavior; A behavior core module (30) is connected to the experience core module (20), and the behavior core module (30) is configured to: Only when the corresponding entity or digital behavior is received from the experience core module (20) and the corresponding task process has been approved by the governance judgment result, the corresponding entity or digital behavior is executed; after the behavior is executed, the behavior result and diagnostic information are returned to the experience core module (20); wherein, the behavior core module (30) does not have the authority to initiate governance judgment, modify governance judgment result, or bypass governance judgment to directly execute behavior; wherein: the cognitive core module (10), the experience core module (20) and the behavior core module (30) are logically distinct from each other and have a forward permission restriction relationship, and the execution of any behavior must be subject to the aforementioned governance judgment result and task process confirmation as a prerequisite.

2. The governance-based autonomous processing system as described in claim 1, wherein, The three states of the governance judgment result correspond to the following processing methods: When it is an allowed state, the task process is allowed to enter the subsequent behavior execution stage; when it is a rejected state, the task process is suspended and no entity or digital behavior is executed; when it is a conditional state, the task process is suspended and the governance judgment is re-performed after supplementary information or verification results are received.

3. The governance-based autonomous processing system as described in claim 1, wherein, When the governance judgment result is conditional, the experience core module (20) is configured to make at least one supplementary information request or verification requirement to human users or external systems, and after obtaining the supplementary information or verification result, incorporate it into the subsequent request parsing or governance judgment process.

4. The governance-based autonomous processing system as described in claim 1, wherein, The experience core module (20) establishes or confirms the corresponding task flow after receiving the governance judgment result output by the cognitive core module (10), and only provides the execution instructions corresponding to the task flow to the behavior core module (30) if the task flow is allowed by the governance judgment result, so that the behavior core module (30) can execute the corresponding entity or digital behavior.

5. The governance-based autonomous processing system as described in claim 1, wherein, After completing the corresponding physical or digital behavior, the behavioral core module (30) provides the behavioral results or diagnostic information to the experience core module (20). Under the condition that the governance conditions are met, the experience core module (20) feeds back the sorted results to the cognitive core module (10) for the cognitive core module (10) to make subsequent governance judgments or update cognitive data.

6. The governance-based autonomous processing system as described in claim 1, wherein, The governance-related data used within the cognitive core module (10) are managed hierarchically according to their nature or governance level, including at least a core governance data level and a non-core data level. The content of the core governance data level is not modified by the task process processing results or sorting results of the experience core module (20), while the content of the non-core data level can be updated based on the sorting results provided by the experience core module (20) if the governance conditions are met.

Images