Business process-oriented data processing method and device, equipment, storage medium

Abstract

This invention relates to the field of artificial intelligence technology and discloses a data processing method, apparatus, device, and storage medium for business processes. The method includes: semantically registering each basic tool to obtain a tool capability catalog; parsing the target business process text to obtain a skill script, where the target business process text contains the business process; compiling the skill script based on the tool capability catalog to obtain a tool call graph; wherein, basic tools serve as nodes in the tool call graph, and the data flow relationships between basic tools serve as edges in the tool call graph; integrating information based on the security attribute information of the nodes to obtain a skill-level capability contract text used to constrain the business process; and executing the business process based on the tool call graph and the skill-level capability contract text. This method can be applied to process processing scenarios in fintech and healthcare, improving the execution efficiency of business processes.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence technology and natural language processing technology, and in particular to a data processing method, apparatus, device, and storage medium oriented towards business processes. Background Technology

[0002] In the field of artificial intelligence technology, related business processes can be completed by invoking multiple tools. For example, in the financial sector, multiple financial tools (databases, computing engines, API interfaces, office software) can be used to complete financial business processes (such as regulatory report generation, risk data extraction, and customer information verification). Similarly, in the medical field, multiple medical tools (databases, laboratory information systems, data anonymization tools) can be used to complete medical business processes (such as patient information verification, analysis and reporting of imaging and laboratory data, and drug management and prescription approval).

[0003] Currently, different tools have different calling methods and data formats, requiring developers to deeply intervene in business processes and write hard-coded scripts to call different tools. However, these scripts are inefficient to build and have poor reusability, resulting in low execution efficiency of business processes. Summary of the Invention

[0004] This invention provides a data processing method, apparatus, computer equipment, and medium for business processes, in order to solve the technical problem of low script construction efficiency and poor script reusability, which in turn leads to low execution efficiency of business processes.

[0005] Firstly, a data processing method oriented towards business processes is provided, the method comprising: At least two basic tools for providing data processing services are acquired, and semantic registration is performed on each of the basic tools to obtain a tool capability directory; wherein, the tool capability directory includes tool interfaces for calling the basic tools and security attribute information for constraining the process of calling the basic tools; Obtain the target business process text, parse the target business process text to obtain the skill script, wherein the target business process text is natural language-based text and contains the business process; The skill scripts are compiled based on the tool capability catalog to obtain a tool call graph; wherein, the basic tools serve as nodes in the tool call graph, and the data flow relationships between the basic tools serve as edges in the tool call graph; Information is integrated based on the security attribute information of the nodes to obtain a skill-level capability contract text used to constrain the business process; The business process is executed based on the tool call graph and the skill-level capability contract text.

[0006] Secondly, a data processing apparatus oriented towards business processes is provided, the apparatus comprising: The tool registration module is used to acquire at least two basic tools for providing data processing services, perform semantic registration on each basic tool, and obtain a tool capability directory; wherein, the tool capability directory includes tool interfaces for calling the basic tools and security attribute information for constraining the process of calling the basic tools; The process parsing module is used to obtain the target business process text, parse the target business process text to obtain the skill script, wherein the target business process text is natural language-based text and contains the business process; The skill compilation module is used to compile the skill scripts based on the tool capability directory to obtain a tool call graph; wherein, the basic tools serve as nodes in the tool call graph, and the data flow relationships between the basic tools serve as edges in the tool call graph; The security constraint module is used to integrate information based on the security attribute information of the node to obtain a skill-level capability contract text used to constrain the business process; The process execution module is used to execute the business process based on the tool call graph and the skill level capability contract text.

[0007] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned business process-oriented data processing method.

[0008] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the aforementioned business process-oriented data processing method.

[0009] In the aforementioned solution implemented by the data processing method, apparatus, computer equipment, and storage medium for business processes, semantic registration can be performed on each basic tool in advance to obtain a tool capability catalog. Each basic tool presents its tool interface and security attribute information in the tool capability catalog, facilitating the invocation of different basic tools. Furthermore, the target business process text can be obtained and parsed into a skill script. Next, the skill script is compiled based on the tool capability catalog to obtain a tool call graph. This allows the skills required to complete the business process to be converted into the tools to be invoked, and provides the tool invocation order. Then, based on the security attribute information corresponding to the nodes in the tool call graph, the skill-level capability contract text corresponding to the entire business process can be integrated to obtain the final business process. Finally, the business process is executed based on the tool call graph and the skill-level capability contract text. In this invention, for business processes that require the invocation of different tools, users can provide target business process text in natural language. The invention cleverly avoids the high complexity of writing scripts to execute business processes due to differences in tool invocation methods and data formats, by first parsing the target business process text into skill scripts, then compiling the skill scripts into tool invocation graphs based on the tool capability catalog, and generating skill-level capability contract texts corresponding to the business process. This significantly reduces the difficulty of business process execution and improves its efficiency. Furthermore, for different business processes, the tool capability catalog can be reused to compile the skill scripts, resulting in a high reusability rate. Compared to existing technologies, this reduces the construction and maintenance costs of scripts used to invoke different tools. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic diagram of an application environment for a data processing method oriented towards business processes according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a data processing method oriented towards business processes according to an embodiment of the present invention; Figure 3 yes Figure 2 A schematic diagram of a specific implementation method for step S10; Figure 4 yes Figure 2 A schematic diagram of a specific implementation method for step S50; Figure 5 yes Figure 4A schematic diagram of a specific implementation method for step S51; Figure 6 This is a schematic diagram of a data processing device oriented towards business processes according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention; Figure 8 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] The data processing method for business processes provided in this invention can be applied to, for example... Figure 1In this application environment, the client communicates with the server via a network. The server can obtain at least two basic tools for providing data processing services, perform semantic registration on each basic tool to obtain a tool capability catalog. The server can receive the target business process text from the client, parse the target business process text to obtain a skill script, compile the skill script based on the tool capability catalog to obtain a tool call graph. The basic tools serve as nodes in the tool call graph, and the data flow relationships between the basic tools serve as edges in the tool call graph. Information is integrated based on the security attribute information of the nodes to obtain the skill-level capability contract text used to constrain the business process. The business process is executed based on the tool call graph and the skill-level capability contract text, and the execution result can be fed back to the client. In this invention, for business processes that require the invocation of different tools, users can provide target business process text in natural language. The invention cleverly avoids the high complexity of writing scripts to execute business processes due to differences in tool invocation methods and data formats, significantly reducing the difficulty and efficiency of business process execution. Furthermore, the tool capability directory can be reused to compile the skill scripts for different business processes, resulting in a high reusability rate. Compared to existing technologies, this reduces the construction and maintenance costs of scripts used to invoke different tools. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a standalone server or a server cluster. Specific embodiments of this invention will be described in detail below.

[0014] Please see Figure 2 As shown, Figure 2 A flowchart illustrating a business process-oriented data processing method provided in an embodiment of the present invention includes the following steps: S10: Obtain at least two basic tools for providing data processing services, perform semantic registration for each basic tool, and obtain a tool capability catalog.

[0015] The data processing method for business processes provided by this invention can be applied to process automation engines in various application scenarios. These engines can be implemented through a server-side component that can register tools. For example, in the financial industry (such as insurance applications), it is common to call tools such as databases, computing engines, APIs, and office software to automate financial processes. The server can register these tools in a tool capability directory for easy and unified access to the required tools later. Similarly, in the healthcare industry (such as drug management and prescription approval), it is common to call tools such as databases, laboratory information systems, and data anonymization tools to automate healthcare business processes. The server can register these tools in a tool capability directory for easy and unified access to the required tools later.

[0016] It should be noted that basic tools can provide certain data processing services. For example, the computing engine provides computing services, the database provides data reading or writing services, and the API interface provides services such as numerical calculation, data cleaning, and data transformation.

[0017] It's important to understand that semantic registration here refers to the process of describing, registering, and aligning the capabilities of basic tools using structured, queryable semantic information. The core idea of ​​semantic registration is to ensure that the server explicitly knows what each basic tool can do, under what conditions it can be used, and how it can be discovered and combined. The tool capability catalog can include the tool interfaces used to call each basic tool, as well as security attribute information used to constrain the process of calling each basic tool. For example, Figure 3 As shown, step S10, which involves semantic registration for each basic tool to obtain a tool capability catalog, includes the following steps: S11: Define the interface for each basic tool to obtain the tool interface; S12: Perform security annotations on each basic tool to obtain security attribute information; among which, the security attribute information includes the minimum permissions required by the tool, data sensitivity, side effects, and potential failure modes; S13: Integrate the tool interfaces and security attribute information of all basic tools to obtain the tool capability catalog.

[0018] For steps S11-S13, the server defines a standardized tool interface for each type of basic tool (such as the SQL query engine, Python numerical computing library, internal data API, and email sending). This tool interface can include: the tool name, function description, input mode, output mode, and invocation method. For example, the tool interface for the database query tool is: query_db(sql:str,timeout:int)-DataFrame, where query_db is the function name, usually an abbreviation for "query database," (sql:str,timeout:int) is the parameter list, sql contains the SQL query text to be executed, str indicates that the data type of the SQL query text is string, timeout represents the query timeout period, usually in seconds, and int indicates that the query timeout period is an integer. Timeout means that if the query is not completed within the specified time, it will be aborted and an error will be thrown. DataFrame refers to returning a data frame, commonly found in data analysis libraries such as Pandas and Polars. A data frame is a data table with a row and column structure that can be used for further processing and analysis.

[0019] The aforementioned security labeling refers to the mandatory requirement to label the security attribute information of the tool in the interface definition. Security attribute information may include: minimum required permissions (such as "read-only access to customer table A"), data sensitivity (such as "output includes PII field"), side effects (such as "will write to audit log"), and potential failure modes (such as "network timeout, SQL syntax error").

[0020] It is important to understand that by centrally managing all registered basic tools and their semantic interfaces and security attribute information, a machine-readable and searchable tool capability catalog can be formed.

[0021] For example, the tool capability catalog includes tool interfaces and security attribute information for tools such as SQL query engines, Python numerical computing libraries, internal data APIs, and email sending, which helps financial business processes call upon the tools in the tool capability catalog to automate processes.

[0022] As can be seen from the above scheme, registering different basic tools to the tool capability directory from two dimensions, namely tool interface and security attribute information, can facilitate the uniformity of subsequent tool calls and reduce the adverse effects of differences between different tools on the execution of business processes.

[0023] S20: Obtain the target business process text, parse the target business process text to obtain the skill script. The target business process text is natural language-based text and contains the business process.

[0024] In step S20, the server designs a declarative domain-specific language that allows users (or agents) to describe complex business processes in a high-level, business-friendly manner. For example, the target process business text is "Generate a monthly sales report for a certain product and send it to the supervisor via email." This text is in natural language, and the business process it contains can be simplified to generating a report and sending it to the supervisor via email.

[0025] After the server obtains the target business process from the client, it parses the target business process text to obtain a skill script. For example, the target business process text "Generate a monthly sales report for a certain product and send it to the supervisor via email" can be parsed to obtain the following skill script: FetchSalesData(product, month)->AnalyzeTrend->GenerateReport->SendEmail(to=manager,attachment=report). Here, FetchSalesData(product, month) represents the sales data set of a certain product for a specified month, usually returned in tabular or data structure form, such as DataFrame, list, or dictionary. This process may involve queries to databases, data warehouses, or external APIs. AnalyzeTrend represents the skill that takes the sales data from the previous step as input, performs trend analysis, and obtains the trend analysis results. The trend analysis results typically include trend direction (upward / downward / stable), growth rate, seasonality, volatility, key indicators, etc., and may still be returned in structured data form (e.g., an object containing indicator fields). This skill may involve tools such as numerical calculation engines. GenerateReport represents the skill that generates a report based on the trend analysis results. This skill may involve tools such as statistical models. SendEmail(to=manager, attachment=report) indicates the skill of sending a generated report to a specified recipient, with the report file as the attachment. This skill may involve email sending tools and data anonymization tools.

[0026] Understandably, in step S20, when parsing the target business process text into a skill script, it is unaware of which tools need to be invoked. The main purpose is to transform the multiple steps involved in the target business process text into multiple skills, facilitating the subsequent determination of the required tools for each skill. For example, the target business process text can be re-described using a skill domain-specific language by a business expert or intelligent agent, thus forming a skill DSL script.

[0027] It should be noted that the aforementioned Domain-Specific Language (DSL) usually refers to a programming or configuration language designed for a specific domain, used to define, combine, and execute tasks, rules, and processes within that domain, rather than a general-purpose programming language.

[0028] S30: Compile skill scripts based on the tool capability catalog to obtain a tool call graph. Basic tools serve as nodes in the tool call graph, and the data flow relationships between basic tools are represented as edges. Data flow relationships include control flow dependencies and data flow dependencies.

[0029] The server-side features an automated skill compiler that parses DSL scripts into tool call graphs. These tool call graphs are directed acyclic graphs (DAGs). Nodes in the tool call graph represent basic tool calls, and edges represent data flow or control flow dependencies. Based on the tool capability catalog, the automated skill compiler automatically decomposes and maps each higher-order step in the skill script to a specific sequence of tool interface calls.

[0030] S40: Integrate information based on the security attribute information of the nodes to obtain the skill-level capability contract text used to constrain business processes.

[0031] A skill-level capability contract text refers to the formal text or structured metadata describing the capabilities, input / output constraints, non-functional attributes, version, governance information, etc., of a skill or a group of skills. The server-side includes a contract synthesizer and validator. After a skill is compiled, the server automatically aggregates the security attribute information of all tools involved in that skill through the contract synthesizer and validator, synthesizing a global skill-level capability contract text. This skill-level capability contract text defines the maximum set of permissions, data flow boundaries, and audit requirements required for skill execution.

[0032] S50: Execute business processes based on tool call graphs and skill-level capability contract texts.

[0033] Specifically, under the constraints of the skill-level capability contract text, relevant tools can be invoked sequentially according to the tool call graph, thereby automating business processes.

[0034] In some embodiments of the invention, such as Figure 4 As shown, step S50, which is the execution of the business process based on the tool call graph and skill-level capability contract text, includes the following steps: S51: Follow the skill level ability contract text and run the tool call nodes in the diagram in sequence; S52: In response to the detection of a runtime failure in a node in the tool call graph, the error detection and recovery strategy of the node is executed; wherein, the error detection and recovery strategy is generated based on the node's security attribute information; S53: In response to the detection that the last node in the tool call graph has run successfully, the business process is determined to be complete.

[0035] In step S51, when skill execution is triggered, the nodes in the tool call diagram are executed sequentially under the constraints of the skill-level ability contract text. For example, Figure 5 As shown, step S51, which involves following the skill-level ability contract text and sequentially running the nodes in the tool call diagram, includes the following steps: S511: When a node in the execution tool call graph is triggered, the execution context of the execution node is obtained, and the execution context is security verified based on the skill-level capability contract text to obtain the verification result; S512: When the verification result is successful, the running tool calls the node in the diagram.

[0036] For steps S511-S512, the execution context (such as executor identity and environment) when running a node must meet the requirements of the skill-level capability contract text before the tool can call the node in the graph to run normally, thereby achieving secure tool invocation. In some embodiments, if the verification result is that the verification fails, the running of the tool call the node in the graph is interrupted.

[0037] In some embodiments of the present invention, step S511, i.e., when the verification result is successful, the tool calls the node in the graph, includes the following steps: when the verification result is successful, a dynamic permission sandbox is started, and within the dynamic permission sandbox, the node is granted the minimum necessary permissions according to the skill-level capability contract text; within the dynamic permission sandbox and based on the minimum necessary permissions, the tool calls the node in the graph. Specifically, when a node is running, it is placed in a dynamic permission sandbox. This sandbox strictly assigns the minimum necessary permissions to the skill running instance according to the skill-level capability contract text (such as only being able to access specific database tables, only being able to send emails to specified mailing lists), and monitors all tool calls to prevent unauthorized operations. In this way, the security of business process execution is improved.

[0038] During the execution of steps S52-S53, some nodes may experience operational failures, such as email sending failure when calling the email sending tool. In such cases, fault recovery needs to be performed according to the error detection and recovery strategy of that node, such as a strategy of "retry 3 times after failure". When the tool calls the last node in the diagram and runs successfully, it indicates that the business process has been completed. At this point, the server can generate and send a notification message to the client indicating that the process has been successfully executed.

[0039] As can be seen from the above scheme, although different basic tools have their own security attribute information, this scheme uses the skill-level capability contract text to constrain the process of calling tools and uses error detection and recovery strategies to achieve automatic fault repair, thereby improving the security and reliability of executing business processes.

[0040] In some embodiments of the invention, before executing the error detection and recovery strategy for a node in response to a detected operational failure in the tool call graph, the method further includes: accessing a preset business strategy library based on the potential failure modes of the nodes in the tool call graph to obtain candidate strategies; and filtering the candidate strategies to obtain error detection and recovery strategies.

[0041] For the above solution, the potential failure modes of a tool refer to various malfunctions, failures, or performance degradations that may occur when a tool is invoked. The preset business strategy library contains failure scenarios and strategies for resolving them. Based on the selection of candidate strategies from the preset business strategies, error detection and recovery strategies can be obtained. For example, for the potential failure mode "database query timeout," the strategy is "retry a maximum of 2 times, with a 5-second interval between each retry"; for the potential failure mode "email sending failure," the strategy is "temporarily store the report in the specified directory and send a system alarm."

[0042] As can be seen from the above scheme, matching error detection and recovery strategies based on the potential failure modes of tools in security attribute information can effectively reduce the risk of tool failure without having to consider other information in security attribute information, thus saving processing resources.

[0043] In some embodiments of this invention, the data processing method for business processes further includes: when the verification result is successful, activating a full-link audit tracer; and recording the process of nodes in the tool call graph based on the full-link audit tracer to obtain audit logs. Specifically, the server may have a full-link audit tracer. This tracer strictly follows the audit requirements in the skill-level capability contract text, recording the complete trace of skill execution, including: start / end time, input / output summary (automatically anonymized) for each tool call, triggered errors and recovery actions, and final result status. The audit logs are immutable and used for post-event review and compliance reporting.

[0044] As can be seen, in the above steps S10 to S50, for business processes that require calling different tools, users can provide target business process text in natural language. The solution cleverly avoids the high complexity of writing scripts to execute business processes due to differences in tool calling methods and data formats, by first parsing the target business process text into skill scripts, then compiling the skill scripts into tool call graphs based on the tool capability catalog, and generating the corresponding skill-level capability contract text for the business process. This significantly reduces the difficulty of business process execution and improves its efficiency. Furthermore, for different business processes, the tool capability catalog can be reused to compile the skill scripts, resulting in a high reusability rate. Compared to existing technologies, this reduces the construction and maintenance costs of scripts used to call different tools.

[0045] The following example illustrates the automatic generation and submission of a daily regulatory report.

[0046] Scenario: A securities company needs to automatically generate a daily trading supervision report after the end of each trading day and submit it to its internal risk control system and a designated email address of the regulatory agency. In this scenario, the data processing method oriented towards the business process may include the following steps: 1. Tool registration to create a tool capability directory: Register Tool A: query_trading_db, permissions: read-only access to the daily transaction details table, security attribute: output includes transaction codes (sensitive). Register Tool B: calculate_risk_metrics, permissions: local calculation, no risk of data leakage. Register Tool C: generate_pdf_report, permissions: use a specified template. Register Tool D: upload_to_internal_risk_system, permissions: upload files to a specific API endpoint. Register Tool E: send_encrypted_email, permissions: send encrypted emails to a pre-defined list of regulatory email addresses.

[0047] 2. Process Analysis: The target business process text, "Automatically generate a daily transaction supervision report after the end of each trading day and submit it to the internal risk control system and the designated email address of the regulatory agency," is analyzed into a skill DSL script: GetTradeData->CalcMetrics->GenReport->UploadToInternal->SendToRegulator. Here, GetTradeData represents extracting / acquiring transaction data; CalcMetrics represents performing indicator calculations and statistical analysis on the acquired transaction data; GenReport represents generating a report based on the results of CalcMetrics and the raw data; UploadToInternal represents uploading the generated report or related data to the internal system or data platform; and SendToRegulator represents sending the report or data from the internal system to the regulatory authority.

[0048] 3. Skill Compilation: The compiler maps skill scripts to specific tool call graphs, such as [A->B->C->D->E]. Here, A to E represent different tools.

[0049] 4. Generate skill-level ability contract text: Synthesis contract requirements: The executing identity must have "transaction data reading", "internal system upload", and "regulatory external release" permissions simultaneously; When the skill is running, the sensitive output of tool A must be processed in the sandbox after being passed to B, and finally encrypted and released by tool E.

[0050] 5. Set the policy: Set "Retry 3 times after failure" for tool D (upload); set "Save the report to secure storage and issue a high-risk alarm immediately after failure" for tool E (send).

[0051] 6. Secure Execution: This skill is automatically triggered daily at 17:00. Before execution, the system verifies that the current service account's permissions meet the contract. The skill operates in a sandbox: Data read by tool A is marked as sensitive and can freely flow to B and C within the sandbox; PDF reports generated by tool C are automatically encrypted and watermarked when sent out of the sandbox to D and E. During execution, if tool D fails to upload on the first attempt, the system automatically retryes according to the policy and succeeds. Tool E successfully sends the encrypted data. The end-to-end audit tracker records in detail the data flow, the time and results of each step, and the retry events of tool D.

[0052] This method transforms a highly sensitive, multi-tool-based routine regulatory process into a secure, reliable, and auditable automated skill. While ensuring absolute compliance, it frees up manpower and eliminates the risks of forgetfulness, delays, or errors that may arise from manual operations.

[0053] Referring to the above solutions, this invention addresses the security, compliance, and reliability challenges faced when combining multiple tools in financial or medical scenarios. Key technical highlights include: providing a unified semantic interface and security attribute description for heterogeneous tools; supporting the visual or scripted combination of basic tools into secure and reliable advanced business skills; and strictly enforcing permissions, data flow, and auditing policies in a sandbox environment. This invention improves the efficiency of skill development in complex business processes such as financial report generation and compliance checks, and reduces operational risks caused by tool misuse, providing a secure and controllable intelligent core engine for financial business automation.

[0054] This invention also offers the following advantages: 1. Enhanced Business Automation Efficiency: Through advanced DSLs and automated synthesis, business personnel can orchestrate complex processes in a near-natural language manner, improving the efficiency of building and deploying complex skills and reducing reliance on specialized development resources. 2. Minimal Convergence of Operational and Compliance Risks: Through capability contracts and dynamic sandboxes, "least privilege" and "full-process monitoring" of complex financial operations are achieved, reducing security incidents caused by tool misuse or unauthorized operations, and ensuring complete and traceable audit records. 3. Significantly Enhanced System Robustness: Automatically generated, semantically aware error recovery strategies enable complex processes to be automatically and gracefully handled or rolled back when some stages fail, improving the overall success rate and reliability of automated tasks. 4. Empowering Intelligent Agent Security Exploration: A framework for secure exploration and tool combination for intelligent agents in the financial and medical fields is provided. Within the security boundaries of DSLs and contracts, intelligent agents can attempt to combine new skills, and their behavior is manageable, controllable, and auditable, laying the foundation for higher-level business automation intelligence.

[0055] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0056] In one embodiment, a business process-oriented data processing apparatus is provided, which corresponds one-to-one with the business process-oriented data processing methods described in the above embodiments. For example... Figure 6 As shown, this business process-oriented data processing device includes a tool registration module 101, a process parsing module 102, a skill compilation module 103, a security constraint module 104, and a process execution module 105. Detailed descriptions of each functional module are as follows: The tool registration module 101 is used to obtain at least two basic tools for providing data processing services, perform semantic registration on each basic tool, and obtain a tool capability directory; wherein, the tool capability directory includes tool interfaces for calling basic tools and security attribute information for constraining the process of calling basic tools; The process parsing module 102 is used to obtain the target business process text, parse the target business process text to obtain the skill script, wherein the target business process text is natural language-based text and contains the business process. The skill compilation module 103 is used to compile skill scripts based on the tool capability catalog to obtain a tool call graph; in which basic tools are used as nodes of the tool call graph, and the data flow relationship between basic tools is used as edges of the tool call graph; The security constraint module 104 is used to integrate information based on the security attribute information of the nodes to obtain the skill-level capability contract text used to constrain business processes. Process execution module 105 is used to execute business processes based on tool call graphs and skill-level capability contract texts.

[0057] In one embodiment, the tool registration module 101 is specifically used for: Define an interface for each basic tool to obtain a tool interface; the tool interface includes the tool name, tool function description, tool input mode, tool output mode, and tool invocation method of the basic tool; Each basic tool is marked with a security label to obtain security attribute information; the security attribute information includes the minimum permissions required by the tool, data sensitivity, side effects, and potential failure modes; The tool capability catalog is obtained by integrating the tool interfaces and security attribute information of all basic tools.

[0058] In one embodiment, the process execution module 105 is specifically used for: Follow the skill-level ability contract text and run the tools to call the nodes in the diagram in sequence; In response to the detection of a runtime failure in a node in the tool call graph, the node's error detection and recovery strategy is executed; wherein, the error detection and recovery strategy is generated based on the node's security attribute information; If the last node in the tool call graph is detected to have run successfully, the business process is considered complete.

[0059] In one embodiment, before executing the node error detection and recovery strategy in response to detecting a runtime failure in a node in the tool call graph, the data processing device for the business process further includes: a strategy setting module, specifically used for: Based on the potential failure modes of the nodes in the tool call graph, the preset business strategy library is accessed to obtain candidate strategies; Candidate strategies are screened to obtain error detection and recovery strategies.

[0060] In one embodiment, the process execution module 105 is specifically used for: When a node in the execution tool call graph is triggered, the execution context of the execution node is obtained, and the execution context is security verified based on the skill-level capability contract text to obtain the verification result. If the verification result is successful, the tool will call the node in the graph.

[0061] In one embodiment, the process execution module 105 is specifically used for: When the verification result is successful, the dynamic permission sandbox is started, and the node is given the minimum necessary permissions according to the skill level capability contract text in the dynamic permission sandbox. Within a dynamic permissions sandbox, nodes in the tool call graph are executed based on the minimum necessary permissions.

[0062] In one embodiment, the data processing apparatus for business processes further includes: an audit trail module, used for: Once the verification result is successful, enable the end-to-end audit tracer. The audit log is obtained by recording the process of calling nodes in the tool's call graph using a full-link audit tracer.

[0063] This invention provides a data processing device for business processes. For business processes that require the invocation of different tools, it supports users providing target business process text in natural language. The device first parses the target business process text into skill scripts, then compiles the skill scripts into tool call graphs based on a tool capability directory, and generates corresponding skill-level capability contract texts for the business process. This cleverly avoids the high complexity of writing scripts to execute business processes due to differences in tool invocation methods and data formats, significantly reducing the difficulty and efficiency of business process execution. Furthermore, for different business processes, the tool capability directory can be reused to compile the skill scripts, resulting in a high reusability rate. Compared to existing technologies, this reduces the construction and maintenance costs of scripts used to invoke different tools.

[0064] Specific limitations regarding the business process-oriented data processing device can be found in the limitations of the business process-oriented data processing method described above, and will not be repeated here. Each module in the aforementioned business process-oriented data processing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0065] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 7As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements a business process-oriented data processing method, server-side functions or steps.

[0066] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 8 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements client-side functions or steps of a business process-oriented data processing method.

[0067] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: Obtain at least two basic tools for providing data processing services, perform semantic registration on each basic tool, and obtain a tool capability catalog; wherein, the tool capability catalog includes tool interfaces for calling basic tools and security attribute information for constraining the process of calling basic tools; Obtain the target business process text, parse the target business process text to obtain the skill script. The target business process text is natural language-based text and contains the business process. The skill scripts are compiled based on the tool capability catalog to obtain a tool call graph; in this graph, basic tools are the nodes and the data flow relationships between basic tools are the edges. Information is integrated based on the security attribute information of the nodes to obtain the skill-level capability contract text used to constrain business processes; Business processes are executed based on tool call graphs and skill-level capability contract texts.

[0068] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Obtain at least two basic tools for providing data processing services, perform semantic registration on each basic tool, and obtain a tool capability catalog; wherein, the tool capability catalog includes tool interfaces for calling basic tools and security attribute information for constraining the process of calling basic tools; Obtain the target business process text, parse the target business process text to obtain the skill script. The target business process text is natural language-based text and contains the business process. The skill scripts are compiled based on the tool capability catalog to obtain a tool call graph; in this graph, basic tools are the nodes and the data flow relationships between basic tools are the edges. Information is integrated based on the security attribute information of the nodes to obtain the skill-level capability contract text used to constrain business processes; Business processes are executed based on tool call graphs and skill-level capability contract texts.

[0069] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0070] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0071] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0072] It should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with the knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0073] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A data processing method oriented towards business processes, characterized in that, The method includes: At least two basic tools for providing data processing services are acquired, and semantic registration is performed on each of the basic tools to obtain a tool capability directory; wherein, the tool capability directory includes tool interfaces for calling the basic tools and security attribute information for constraining the process of calling the basic tools; Obtain the target business process text, parse the target business process text to obtain the skill script, wherein the target business process text is natural language-based text and contains the business process; The skill scripts are compiled based on the tool capability catalog to obtain a tool call graph; wherein, the basic tools serve as nodes in the tool call graph, and the data flow relationships between the basic tools serve as edges in the tool call graph; Information is integrated based on the security attribute information of the nodes to obtain a skill-level capability contract text used to constrain the business process; The business process is executed based on the tool call graph and the skill-level capability contract text.

2. The method according to claim 1, characterized in that, The semantic registration of each of the aforementioned basic tools to obtain a tool capability catalog includes: For each of the basic tools, an interface is defined to obtain a tool interface; wherein, the tool interface includes the tool name, tool function description, tool input mode, tool output mode, and tool invocation method of the basic tool; Each of the aforementioned basic tools is marked with security information to obtain security attribute information; wherein, the security attribute information includes the minimum permissions required by the tool, data sensitivity, side effects, and potential failure modes; The tool capability catalog is obtained by integrating the tool interfaces and security attribute information of all the basic tools.

3. The method according to claim 2, characterized in that, The execution of the business process based on the tool call graph and the skill-level capability contract text includes: Follow the skill level ability contract text and run the nodes in the tool call diagram in sequence; In response to the detection of a runtime failure in a node in the tool call graph, an error detection and recovery strategy for that node is executed; wherein, the error detection and recovery strategy is generated based on the node's security attribute information; Upon detecting that the last node in the tool call graph has been successfully executed, the business process is determined to be complete.

4. The method according to claim 3, characterized in that, Before executing the error detection and recovery strategy for a node in the tool call graph in response to detecting a runtime failure, the method further includes: Based on the potential failure modes of the nodes in the tool call diagram, the preset business strategy library is accessed to obtain candidate strategies; The candidate strategies are filtered to obtain the error detection and recovery strategy.

5. The method according to claim 3, characterized in that, The process of following the skill level ability contract text and sequentially running the nodes in the tool call graph includes: When a node in the tool call graph is triggered, the execution context of the node is obtained, and the execution context is security verified based on the skill-level capability contract text to obtain the verification result. If the verification result is successful, the tool will be run to call the node in the graph.

6. The method according to claim 5, characterized in that, When the verification result is successful, the tool is run to call the nodes in the graph, including: When the verification result is successful, the dynamic permission sandbox is activated, and the node is granted the minimum necessary permissions according to the skill level capability contract text in the dynamic permission sandbox. The nodes of the tool call graph are run within the dynamic permission sandbox and based on the minimum necessary permissions.

7. The method according to claim 5, characterized in that, The method further includes: If the verification result is successful, the end-to-end audit tracer will be activated. The process of running the tool in the call graph is recorded by the full-link audit tracer to obtain the audit log.

8. A data processing device oriented towards business processes, characterized in that, The device includes: The tool registration module is used to acquire at least two basic tools for providing data processing services, perform semantic registration on each basic tool, and obtain a tool capability directory; wherein, the tool capability directory includes tool interfaces for calling the basic tools and security attribute information for constraining the process of calling the basic tools; The process parsing module is used to obtain the target business process text, parse the target business process text to obtain the skill script, wherein the target business process text is natural language-based text and contains the business process; The skill compilation module is used to compile the skill scripts based on the tool capability directory to obtain a tool call graph; wherein, the basic tools serve as nodes in the tool call graph, and the data flow relationships between the basic tools serve as edges in the tool call graph; The security constraint module is used to integrate information based on the security attribute information of the node to obtain a skill-level capability contract text used to constrain the business process; The process execution module is used to execute the business process based on the tool call graph and the skill level capability contract text.

9. A computer device, characterized in that, The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images