A management system for the integrated governance and precise application output of data assets across the entire domain.

Abstract

This invention discloses a management system for the integrated governance and precise application output of full-domain data assets, relating to the field of data asset management technology. This management system includes a governance status package generation module, a task constraint package generation module, a dual-state evidence splitting module, a constraint mirror graph generation module, an evidence reversal identification module, a conflict evidence transmission module, an output path reconstruction module, and a responsibility write-back update module. Around the full-domain data asset records and application output tasks, it completes governance status extraction, evidence splitting, graph relationship construction, reversal factor identification, cross-asset transmission, path reconstruction, and feedback updates. This invention transmits governance evidence to the asset call path through supporting evidence surfaces, restricting evidence surfaces, constraint mirror graphs, and same-asset evidence reversal factors, and updates the application output path reconstruction package in conjunction with path feedback data, reducing the disconnect between governance results and application output paths.

Description

Technical Field

[0001] This invention relates to the field of data asset management technology, specifically a management system for the integrated governance and precise application output of data assets across the entire domain. Background Technology

[0002] As organizations such as enterprises, industrial parks, and manufacturing continue to develop their business systems, data such as business processing data, equipment operation data, interface exchange data, document data, indicator statistics data, and collaborative access data are gradually forming data assets covering multiple business domains. Different data assets have different management statuses in terms of source ownership, field definitions, update cycles, quality rules, permission boundaries, and call records. The way data assets are used has also gradually expanded from single queries and fixed reports to tasks such as indicator generation, business tag output, business prompts, and data services.

[0003] Existing data asset management systems typically govern data assets through data catalogs, metadata registration, quality rule verification, tag classification, interface publishing, and report configuration. They can record the basic attributes and quality status of data assets and complete fixed outputs according to manually selected data tables, interfaces, or indicator templates. However, in the scenario of integrated governance of data assets across the entire domain, data assets from different business domains differ in terms of field definitions, update times, quality hit results, permission boundaries, source ownership, and historical call feedback. The output task of the same application will be subject to the joint constraints of multiple governance evidences.

[0004] The limitations of existing technologies include at least the following problems: Existing systems mostly express data asset governance results as catalog items, tag items, quality scores, interface status, or permission boundary markers. These results mainly serve asset registration and manual retrieval, and are difficult to reorganize into executable output paths according to application output tasks. When data assets are called in combination across business domains, situations such as consistent field definitions but different update times, or high quality scores but restricted permission boundaries are likely to occur. Existing catalog retrieval or quality scoring methods are not easy to identify combination conflicts caused by the mutual obscuring of multiple governance constraints, nor are they easy to continuously transmit conflict evidence to asset combination, output definitions, call order, and feedback write-back. As a result, after the data assets of the entire domain are aggregated and governed, the application output path is still disconnected from the governance results. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a management system for the integrated governance and precise application output of cross-domain data assets, which solves the problem of difficulty in transmitting cross-domain data asset governance evidence to the application output path.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a management system for the integrated governance and precise application output of full-domain data assets, comprising: a governance status package generation module, used to extract governance status elements and evidence data based on full-domain data asset records, and generate a full-domain data asset governance status package; a task constraint package generation module, used to extract task basic elements and task constraint elements based on application output tasks, and generate an application output task constraint package; a dual-state evidence splitting module, used to split the full-domain data asset governance status package into supporting evidence surfaces and restrictive evidence surfaces based on the full-domain data asset governance status package and the application output task constraint package; and a constraint mirror diagram generation module, used to construct an original governance dimension relationship diagram based on the supporting evidence surfaces, restrictive evidence surfaces, and application output task constraint packages, and to separate cross-dimensional nodes. The state combination is recombined into a constraint mirror diagram; the evidence reversal identification module is used to identify the location in the constraint mirror diagram where the supporting evidence surface of the same data asset is blocked by the restricted evidence surface, and generate the same-asset evidence reversal factor; the conflict evidence transmission module is used to perform cross-asset transmission in the constraint mirror diagram based on the same-asset evidence reversal factor, and generate a cross-asset mutually exclusive transmission chain; the output path reconstruction module is used to reconstruct the data asset call path corresponding to the application output task based on the same-asset evidence reversal factor and the cross-asset mutually exclusive transmission chain, and generate the application output path reconstruction package; the responsibility write-back update module is used to update the global data asset governance state package, the same-asset evidence reversal factor, the cross-asset mutually exclusive transmission chain and the application output path reconstruction package based on the application output results, call feedback and exception records corresponding to the application output path reconstruction package.

[0007] Furthermore, the steps for generating the global data asset governance status package are as follows: Read global data asset records, which include business domain data records, device-side data records, interface-side data records, document-side data records, indicator-side data records, and collaborative access data records; merge asset objects from the global data asset records to generate data asset objects; extract governance status elements from the data asset objects, including source ownership, field definitions, update cycle, quality rules, permission boundaries, and call records; perform computer vision recognition on the document-side data records to generate evidence data, including layout area evidence, table area evidence, and image field evidence; and encapsulate the governance status elements and evidence data into a global data asset governance status package.

[0008] Furthermore, the steps of computer vision recognition are as follows: read the image file from the document-side data record; perform page area segmentation on the image file to obtain the page area; extract table field information and image field information from the page area; extract field structure evidence from the table field information; extract image object evidence from the image field information; and write the field structure evidence and image object evidence into the evidence data.

[0009] Furthermore, the steps for separating the supporting and restrictive evidence surfaces are as follows: read the governance status elements and evidence data from the global data asset governance status package; read the task constraint elements from the application output task constraint package; write the governance status elements and evidence data that satisfy the task constraint elements into the supporting evidence surface; write the governance status elements and evidence data that block the task constraint elements into the restrictive evidence surface; bind the supporting and restrictive evidence surfaces to the same data asset object, and generate a bi-state evidence index based on the binding result.

[0010] Furthermore, the steps for generating the application output task constraint package are as follows: parse the application output task to obtain the basic task elements, which include the application scenario, output object, and output format; generate task constraint elements based on the basic task elements, which include indicator definitions, field ranges, permission boundary requirements, call constraints, timeliness requirements, and quality requirements; and encapsulate the basic task elements and task constraint elements into the application output task constraint package.

[0011] Furthermore, the steps for generating the constraint mirror graph are as follows: Establish dimensional nodes for the supporting and limiting evidence surfaces according to governance state elements and evidence data; establish task nodes for the application output task constraint package according to task basic elements and task constraint elements; establish an original governance dimensional relationship graph between the dimensional nodes and task nodes; extract the node state combinations that connect across dimensions in the original governance dimensional relationship graph to generate mirror nodes; generate mirror edges according to the supporting and blocking relationships between mirror nodes, and generate the constraint mirror graph based on the mirror nodes and mirror edges.

[0012] Furthermore, the steps for generating the same asset evidence reversal factor are as follows: Read the supporting evidence surface, limiting evidence surface, and application output task constraint package corresponding to the same data asset object in the constraint mirror diagram; find the supporting path where the supporting evidence surface hits the application output task constraint package along the mirror edge; find the blocking path where the limiting evidence surface hits the application output task constraint package along the mirror edge; mark the evidence reversal position when the supporting path and the blocking path both point to the same application output task; generate the same asset evidence reversal factor based on the evidence reversal position, supporting path, and blocking path.

[0013] Furthermore, the steps for generating the cross-asset mutually exclusive transmission chain are as follows: Read the data asset objects matching the task constraint elements in the constraint mirror diagram to generate a candidate asset group; read the data asset objects corresponding to the same asset evidence inversion factor; extract governance status elements from the candidate asset group; generate mutually exclusive transmission sub-chains based on the overlapping states between governance status elements. These sub-chains characterize the constraint mutually exclusive transmission relationship between different data asset objects, and include time-limited mutually exclusive chains, invocation mutually exclusive chains, and feedback mutually exclusive chains; map the mutually exclusive transmission sub-chains to the data asset objects in the candidate asset group; and generate the cross-asset mutually exclusive transmission chain based on the mapped sub-chains.

[0014] Furthermore, the steps for generating the application output path reconstruction package are as follows: Read the application output task constraint package, the same-asset evidence inversion factor, and the cross-asset mutual exclusion transmission chain; generate path type identifiers based on the same-asset evidence inversion factor and the cross-asset mutual exclusion transmission chain; classify data asset objects by call type according to the path type identifiers to obtain call asset groups, which represent the call role of data asset objects in the application output path; generate the main call path, delayed verification path, exception blocking path, and replacement call path according to the call asset groups; generate the application output path reconstruction package based on the main call path, delayed verification path, exception blocking path, and replacement call path; and write the path type identifiers into the application output path reconstruction package.

[0015] Furthermore, the processing steps of the responsibility write-back update module are as follows: Collect path feedback data corresponding to the application output path reconstruction package. Path feedback data includes application output results, call feedback, and exception records. Locate the data asset object corresponding to the path feedback data based on the application output path reconstruction package. Read the corresponding global data asset governance status package, same-asset evidence reversal factor, and cross-asset mutual exclusion transmission chain based on the data asset object. Write the path feedback data into the supporting evidence surface and restrictive evidence surface in the global data asset governance status package. Update the same-asset evidence reversal factor based on the written supporting evidence surface and restrictive evidence surface. Update the cross-asset mutual exclusion transmission chain and application output path reconstruction package based on the updated same-asset evidence reversal factor.

[0016] The present invention has the following beneficial effects: (1) The management system applied to the integrated governance and precise application output of data assets in the whole domain merges the business domain data records, device side data records, interface side data records, document side data records, indicator side data records and collaborative access data records into asset objects through the governance status package generation module, and extracts the source ownership, field definition, update cycle, quality rules, permission boundaries and call records, so that data records from different sources can be classified into the same data asset object; at the same time, the evidence data takes over the page area evidence, table area evidence and image field evidence in the document side data records, so that the field information in the document materials, interface records and indicator records can enter the governance link together with the data asset object, reducing the manual verification burden caused by the scattered storage of field definition, update time and call records.

[0017] (2) The management system applied to the integrated governance and precise application output of the whole domain data assets splits the whole domain data asset governance status package into supporting evidence surface and restrictive evidence surface through the dual-state evidence splitting module, and the constraint mirror diagram generation module writes the supporting relationship and blocking relationship into the constraint mirror diagram, so that situations such as field caliber meeting the task requirements but update time exceeding the deadline, quality rules passing but permission boundaries being restricted can be retained; the evidence reversal identification module searches for evidence reversal positions in the same application output task along the supporting path and blocking path, and generates the same asset evidence reversal factor, so that positive evidence and reverse evidence within the same data asset can be distinguished and recorded, reducing the combination conflict omissions caused by relying solely on directory items, tag items or quality scores for judgment.

[0018] (3) The management system applied to the integrated governance and precise application output of data assets in the whole domain maps the same asset evidence inversion factor to the candidate asset group through the conflict evidence transmission module, and generates a cross-asset mutual exclusion transmission chain, so that the blocking evidence inside a single data asset can be transmitted to the asset combination calling process; the output path reconstruction module generates the main calling path, delayed verification path, abnormal blocking path and replacement calling path according to the same asset evidence inversion factor and the cross-asset mutual exclusion transmission chain, so that the application output task can reorganize the calling order according to the governance evidence; the responsibility write-back update module writes the application output results, call feedback and abnormal records back to the supporting evidence surface and the limiting evidence surface, so that data asset governance, application output and feedback records form a continuous processing link.

[0019] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0020] Figure 1 This is a block diagram of a management system for the integrated governance and precise application output of data assets across the entire domain, as described in this invention.

[0021] Figure 2This is a flowchart illustrating the steps involved in generating a state package for the governance of data assets across the entire domain, as described in this invention, within a management system for the integrated governance and precise application output of data assets across the entire domain. Detailed Implementation

[0022] Please see Figure 1 This invention provides a technical solution: a management system for the integrated governance and precise application output of full-domain data assets, comprising: a governance status package generation module, used to extract governance status elements and evidence data based on full-domain data asset records, and generate a full-domain data asset governance status package; a task constraint package generation module, used to extract task basic elements and task constraint elements based on application output tasks, and generate an application output task constraint package; a dual-state evidence splitting module, used to split the full-domain data asset governance status package into supporting evidence surfaces and restrictive evidence surfaces based on the full-domain data asset governance status package and the application output task constraint package; and a constraint mirror diagram generation module, used to construct an original governance dimension relationship diagram based on the supporting evidence surfaces, restrictive evidence surfaces, and application output task constraint packages, and to combine cross-dimensional node states. The system is reorganized into a constraint mirror diagram; the evidence reversal identification module is used to identify the locations in the constraint mirror diagram where the supporting evidence surface of the same data asset is blocked by the restricted evidence surface, and generate the same-asset evidence reversal factor; the conflict evidence transmission module is used to perform cross-asset transmission in the constraint mirror diagram based on the same-asset evidence reversal factor, and generate a cross-asset mutually exclusive transmission chain; the output path reconstruction module is used to reconstruct the data asset call path corresponding to the application output task based on the same-asset evidence reversal factor and the cross-asset mutually exclusive transmission chain, and generate the application output path reconstruction package; the responsibility write-back update module is used to update the global data asset governance status package, the same-asset evidence reversal factor, the cross-asset mutually exclusive transmission chain, and the application output path reconstruction package based on the application output results, call feedback, and exception records corresponding to the application output path reconstruction package.

[0023] The modules are executed in the following order: data asset record access, application output task parsing, dual-state evidence splitting, constraint mirroring map building, same-asset evidence inversion identification, cross-asset mutual exclusion transmission, output path reconstruction, and responsibility write-back update. The global data asset governance status package serves as the input on the data asset side, the application output task constraint package serves as the input on the task side, the supporting evidence surface and the limiting evidence surface serve as intermediate evidence carriers, and the constraint mirror diagram is used to carry the supporting relationship and the blocking relationship between the data asset governance evidence and the application output task constraints.

[0024] Specifically, such as Figure 2 As shown, the steps for generating the global data asset governance status package are as follows: Read the entire domain data asset records. These records include business domain data records, device-side data records, interface-side data records, document-side data records, metric-side data records, and collaborative access data records. Specifically: Data records are accessed from business systems, equipment acquisition systems, interface exchange systems, document management systems, indicator management systems, and collaborative access systems. Asset source identifier, business domain identifier, data object name, field set, update time, calling method, interface address, document page number, indicator name, and access party identifier are read and grouped according to the record source into business domain data records, device-side data records, interface-side data records, document-side data records, indicator-side data records, and collaborative access data records. In one implementation, the business domain data record includes business processing records and basic records of business objects; the device-side data record includes device operating status records and device collection time records; the interface-side data record includes interface call address, interface return fields, and interface call results; the document-side data record includes document images, scanned pages, and layout pages; the indicator-side data record includes indicator name, indicator definition, and indicator results; and the collaborative access data record includes access party identifier, access data name, and access time.

[0025] The entire domain data asset records are merged into asset objects to generate data asset objects, specifically as follows: Using asset source identifier, business domain identifier, data object name, field set, interface call identifier, and document object identifier as the merging criteria, data records pointing to the same business object, device object, interface object, document object, or indicator object are merged, assigned a unified data asset object identifier, and a mapping relationship is established between the data asset object identifier and the original data record. In one implementation, when the device number in the business domain data record, the acquisition device number in the device-side data record, and the interface return device number in the interface-side data record are consistent, the corresponding records are merged into the same data asset object.

[0026] Extract the governance status elements of data asset objects. These elements include source ownership, field definition, update cycle, quality rules, permission boundaries, and access records. Specifically: Extract the data source unit, data generation system, collection responsibility party, and access responsibility party from the data records corresponding to the data asset object to form the source ownership; Extract field names, field meanings, field units, statistical definitions, and value ranges to form field definitions; Extract the data generation time, data update time, planned update frequency, and most recent update time to form the update cycle; Extract missing check rules, duplicate check rules, range check rules, and correlation check rules to form quality rules; Extract the calling role, calling scope, output object, and usage restrictions to form permission boundaries; extract historical calling time, calling task, calling result, and calling exception to form a calling record.

[0027] Computer vision recognition is performed on document-side data records to generate evidentiary data, which includes layout area evidence, table area evidence, and image field evidence, specifically: Read the image files, scanned files, or layout files corresponding to the document side data records, perform grayscale processing, boundary detection, and region recognition on the document pages, extract the title area, body text area, table area, image area, and annotation area in the page, and generate evidence data based on page number, area type, area location, table title, table header field, row and column structure, cell content, image object, and image object association field.

[0028] The governance status elements and evidence data are encapsulated into a global data asset governance status package, which is as follows: Using the data asset object identifier as the main index, the source ownership, field definition, update cycle, quality rules, permission boundaries, and call records are written into the governance status element set. Page area evidence, table area evidence, and image field evidence are written into the evidence data set. The governance status element set and the evidence data set are then bound to the same data asset object.

[0029] The steps of computer vision recognition are as follows: Reading image files from document-side data records specifically involves: Read the corresponding page image based on the file identifier, page number identifier, and storage path in the document-side data record, and record the file source, page number, image size, image orientation, and associated data asset object identifier of the page image.

[0030] Perform page layout segmentation on the image file to obtain page layout regions, specifically as follows: The system performs grayscale processing, noise removal, skew correction, and boundary detection on the page image, identifies the title area, paragraph area, table area, image area, signature area, and header / footer area on the page, and generates area number, area type, area coordinates, and page number for each page area.

[0031] Extracting table field information and image field information from the layout area, specifically as follows: Identify the table header, row and column lines, cell boundaries, and cell text within the table area to obtain table field information; In the image area, annotation area, and mixed image and text area, identify image objects, image object locations, image object annotations, and image object association fields to obtain image field information; In one implementation, the image object includes a device image, a flowchart node, a statistical graph, or a label symbol, and the associated fields of the image object include field names near the image, legend text, or table numbers.

[0032] Extracting field structure evidence from table field information specifically involves: Read the header fields, field levels, field units, field values, row and column positions, and business meanings of the fields from the table field information, and organize the field names, field levels, field units, field positions, and field value relationships into field structure evidence.

[0033] Extracting image object evidence from image field information specifically involves: Read the image object category, object location, object annotation, correspondence between objects and fields, and the page to which the object belongs from the image field information, and organize the image object category, image object location, image object annotation, and object association fields into image object evidence.

[0034] The field structure evidence and image object evidence are written into the evidence data, specifically as follows: Write the field structure evidence into the table area evidence, write the image object evidence into the image field evidence, and establish a page and area correspondence between the table area evidence, the image field evidence, and the layout area evidence.

[0035] In this implementation plan, the source ownership, field definitions, update cycle, quality rules, permission boundaries and call records of data assets are carried by the governance status elements, and the layout, table and image field information in the document-side data records are carried by the evidence data, so that structured records, interface records, indicator records and document-side records can be classified into the same data asset object, reducing the evidence gap caused by the separation of document-side data and structured data.

[0036] Specifically, the steps for separating supporting and limiting evidence are as follows: Read the governance status elements and evidence data from the global data asset governance status package, specifically: Based on the data asset object identifier, the corresponding global data asset governance status package is read, and the source ownership, field definition, update cycle, quality rules, permission boundaries, call records, page area evidence, table area evidence, and image field evidence are retrieved.

[0037] Read the task constraint elements from the application's output task constraint package, specifically: Read the application output task constraint package based on the application output task identifier, retrieve the indicator scope, field range, permission boundary requirements, call constraints, timeliness requirements and quality requirements, and establish a matching relationship between the task constraint elements and the candidate data asset objects.

[0038] The governance status elements and evidentiary data that satisfy the task constraints are written into the supporting evidence surface, specifically as follows: The governance status elements and evidence data are matched with the task constraint elements one by one. When the field definition conforms to the indicator definition, the field range covers the task fields, the update cycle meets the timeliness requirements, the quality rule is not hit, the permission boundary meets the calling constraints, and the evidence data can support the identification of the task fields, the corresponding governance status elements and evidence data are written into the supporting evidence surface. In one implementation, when the field definition of a certain indicator data asset is consistent with the indicator definition of the application output task, and the field range covers the fields required by the task, the evidence corresponding to the field definition and field range is written into the supporting evidence surface.

[0039] The governance status elements and evidentiary data of the blocking task constraint elements are written into the limiting evidence surface, specifically as follows: The governance status elements and evidence data are compared with the task constraint elements one by one. When the field definition does not match the indicator definition, the field range is missing, the update cycle exceeds the timeliness requirement, the quality rule hit is abnormal, the permission boundary does not meet the call constraint, or the evidence data cannot support the identification of the task field, the corresponding governance status elements and evidence data are written into the restricted evidence surface. In one implementation, if the field definition of a certain indicator data asset meets the indicator definition requirements of the application output task, but the update time of the indicator data asset exceeds the task timeliness requirement, the governance status element corresponding to the update time will be written into the restricted evidence surface.

[0040] The supporting and limiting evidence are bound to the same data asset object, and a bi-state evidence index is generated based on the binding result. Specifically: Using the data asset object identifier as the binding primary key, a corresponding relationship is established between the supporting evidence surface and the restrictive evidence surface corresponding to the same data asset object. An evidence surface identifier, evidence source identifier, task constraint element identifier, and evidence status identifier are generated for each supporting evidence and restrictive evidence, forming a dual-state evidence index.

[0041] In this implementation plan, by writing supporting evidence and limiting evidence under the same data asset object into the supporting evidence surface and the limiting evidence surface respectively, the data asset can be recorded as satisfying a certain task constraint or as being blocked by another task constraint, thus preserving the positive and negative evidence within the same data asset and facilitating the identification of evidence reversal factors within the same asset.

[0042] Specifically, the steps for generating the application output task constraint package are as follows: Analyzing the application's output task yields the basic elements of the task, which include the application scenario, the output object, and the output format, specifically: Read the task name, task source, task purpose, receiving object, output carrier, and output triggering method from the application output task. Determine the application scenario based on the task purpose, determine the output object based on the receiving object, and determine the output format based on the output carrier and output triggering method. Write the application scenario, output object, and output format into the basic elements of the task.

[0043] Task constraint elements are generated based on the basic task elements. These constraints include indicator definitions, field ranges, permission boundary requirements, invocation constraints, timeliness requirements, and quality requirements. The definition and quality requirements of the indicators are determined according to the application scenario, the permission boundary requirements are determined according to the output object, the field range and calling constraints are determined according to the output form, and the timeliness requirements are determined according to the task trigger time and business processing cycle. The definition, field range, permission boundary requirements, calling constraints, timeliness requirements and quality requirements are written into the task constraint elements. In one implementation, when the application output task is to generate monthly indicator results for a certain business domain, the basic elements of the task include the monthly indicator scenario, internal business objects, and the output format of the indicator table. The task constraint elements include the indicator statistical scope, the required field range, the internal call permission boundary, the interface call constraint, the monthly update timeliness requirement, and the quality rule pass requirement.

[0044] The basic elements and constraints of the task are encapsulated into an application output task constraint package, which is as follows: Using the application output task identifier as the main index, the application scenario, output object, and output form are written into the task basic element set, and the indicator scope, field range, permission boundary requirements, call constraints, timeliness requirements, and quality requirements are written into the task constraint element set. The task basic element set and the task constraint element set are then bound together to generate the application output task constraint package.

[0045] In this implementation plan, by splitting application output tasks into basic task elements and task constraint elements, the application output tasks are transformed from task names or templates into constraint carriers that can be matched item by item with the data asset governance status package, so that the data asset call path can be identified and reconstructed according to the task requirements.

[0046] Specifically, the steps for generating the constraint mirror diagram are as follows: The supporting and limiting evidence aspects will be established as dimensional nodes based on governance status elements and evidence data, specifically as follows: Based on the source ownership, field definition, update cycle, quality rules, permission boundaries, call records and evidence data as the dimensional division criteria, support dimension nodes are established for supporting evidence in the supporting evidence plane, and restriction dimension nodes are established for restriction evidence in the restriction evidence plane. The dimension nodes are then bound to the data asset object identifier, evidence plane identifier and evidence source identifier.

[0047] The application output task constraint package will be used to create task nodes based on the basic task elements and task constraint elements, specifically as follows: The task division is based on application scenarios, output objects, output formats, indicator definitions, field ranges, permission boundary requirements, calling constraints, timeliness requirements, and quality requirements. Corresponding task nodes are established and bound to application output task identifiers and task constraint element identifiers.

[0048] Establish the original governance dimension relationship graph between dimension nodes and task nodes, specifically as follows: Dimension nodes are matched with task nodes. When the governance status elements or evidence data corresponding to the dimension node meet the task constraint elements corresponding to the task node, a support relationship edge is established between the dimension node and the task node. When the governance status elements or evidence data corresponding to a dimension node block the task constraint elements corresponding to a task node, an blocking relationship edge is established between the dimension node and the task node, and the supporting relationship edge and the blocking relationship edge are written into the original governance dimension relationship graph.

[0049] Extract the node state combinations of cross-dimensional connections from the original governance dimension relationship graph to generate mirror nodes, specifically as follows: Read the dimension nodes that are connected to task nodes under the same data asset object in the original governance dimension relationship diagram, extract the node state combinations between field caliber and timeliness requirements, quality rules and permission boundary requirements, call records and call constraints, evidence data and field ranges, and reorganize the node state combinations into mirror nodes. In one implementation, when the field caliber dimension node of the same data asset object forms a supporting relationship edge and the update cycle dimension node forms a blocking relationship edge, the state of the field caliber dimension node, the update cycle dimension node and the task node connected to them are combined into a mirror node.

[0050] Mirror edges are generated based on the support and blocking relationships between mirror nodes, and a constrained mirror graph is generated based on the mirror nodes and mirror edges, specifically as follows: Mirror edges are established based on the support and blocking states contained in the mirror nodes. Support mirror edges are established when there are common task constraints between mirror nodes and the support states can be continuously transmitted. Blocking mirror edges are established when there are common task constraints between mirror nodes and the blocking states can be transmitted. A constraint mirror graph is generated from the mirror nodes, support mirror edges, and blocking mirror edges.

[0051] In this implementation plan, governance evidence scattered across field definitions, update cycles, quality rules, permission boundaries, call records, and evidence data is mapped to the same application output task through a constraint mirror graph. This allows the support and blocking states within the same data asset object to be processed within the same graph relationship, reducing conflicts and omissions caused by relying solely on directory items, tag items, or quality scores for judgment.

[0052] Specifically, the steps for generating the asset-backed evidence reversal factor are as follows: Read the supporting evidence surface, limiting evidence surface, and application output task constraint package corresponding to the same data asset object in the constraint mirror diagram, specifically: Based on the data asset object identifier, locate the corresponding supporting dimension node, limiting dimension node, supporting mirror edge, and blocking mirror edge in the constraint mirror graph, and read the application output task constraint package associated with the data asset object.

[0053] The supporting path for finding supporting evidence surfaces along the mirror edges and matching the supporting constraints of the applied output task package is as follows: Starting from the supporting dimension node corresponding to the supporting evidence surface, search along the supporting relationship edge and the supporting mirror edge to find the path that can hit the task constraint element, and record the starting point of the path, the mirror node passed by the path, the ending point of the path and the hit task constraint element as the supporting path.

[0054] The blocking path for finding the constraint evidence surface that hits the application output task constraint package along the mirror edge is as follows: Starting from the constraint dimension node corresponding to the constraint evidence surface, find the path of the blocking task constraint element along the blocking relation edge and the blocking mirror edge, and record the starting point of the path, the mirror node passed by the path, the ending point of the path and the blocked task constraint element as the blocking path.

[0055] When both the supporting path and the blocking path point to the same application output task, mark the evidence reversal position, specifically as follows: By comparing the application output task identifiers corresponding to the supporting path and the blocking path, if the supporting path and the blocking path of the same data asset object both point to the same application output task, and the task constraint element hit by the supporting path is restricted by the task constraint element corresponding to the blocking path, the intersection node of the supporting path and the blocking path is marked as the evidence reversal position. In one implementation, if the field range of a certain interface data asset satisfies the application output task, but the permission boundary of the interface data asset does not meet the output object requirements, then the field range forms a supporting path and the permission boundary forms a blocking path. When both paths point to the same application output task, the intersection node of the field range supporting path and the permission boundary blocking path is marked as the evidence reversal position.

[0056] Based on the evidence reversal location, supporting path, and blocking path, an evidence reversal factor for the same asset is generated, specifically as follows: Using the data asset object identifier and the application output task identifier as the primary key, record the evidence reversal location, supporting path identifier, blocking path identifier, supported task constraint elements, blocked task constraint elements, and reversal type, and encapsulate the above records into the same asset evidence reversal factor.

[0057] In this implementation plan, the location where the supporting path of the same data asset object is blocked by the reverse restriction of the blocked path is recorded by the same asset evidence reversal factor, so that the data asset is no longer simply judged as available or unavailable. It can express the intermediate state that is available but blocked by another governance constraint, which facilitates cross-asset mutual exclusion transmission and output path reconstruction.

[0058] Specifically, the steps for generating a cross-asset mutually exclusive transmission chain are as follows: Read the data asset objects that match the task constraint elements in the constraint mirror diagram, and generate a candidate asset group, specifically as follows: Based on the task constraint elements in the application output task constraint package, search for data asset objects that can satisfy the task constraint elements in the constraint mirror diagram, and group the found data asset objects according to the application output task identifier to generate candidate asset groups.

[0059] Read the data asset object corresponding to the asset evidence reversal factor, specifically: Based on the data asset object identifier in the same asset evidence reversal factor, locate the data asset object that has undergone evidence reversal in the candidate asset group, and read the evidence reversal position, supporting path and blocking path corresponding to the data asset object.

[0060] The governance status elements are extracted from the candidate asset group, specifically as follows: Read the source ownership, field definition, update cycle, quality rules, permission boundaries and call records of each data asset object in the candidate asset group, and establish a candidate asset governance status table according to the data asset object identifier.

[0061] Mutually exclusive propagation subchains are generated based on the overlapping states between governance status elements. These subchains represent the mutually exclusive propagation relationships of constraints between different data asset objects. The mutually exclusive propagation subchains include time-limited mutually exclusive chains, invocation mutually exclusive chains, and feedback mutually exclusive chains, specifically as follows: Compare the update cycles, call records and historical call feedback of different data asset objects within the candidate asset group. When data asset objects correspond to the same task field but have different update time windows, generate a time-limited mutual exclusion chain. When data asset objects require the same interface resource, the same calling permission, or the same output channel, a call mutex chain is generated. When there are conflicting exception records, replacement records, or blocking records in the historical call feedback, a feedback mutual exclusion chain is generated; In one implementation, if two data asset objects can provide the same indicator field, but one data asset object's update time meets the task's timeliness requirement while the other data asset object's update time exceeds the task's timeliness requirement, a timeliness mutual exclusion chain is generated between the two data asset objects.

[0062] The mutex propagation subchain is mapped to the data asset objects in the candidate asset group, specifically as follows: Bind the time-limited mutex chain, the call mutex chain, and the feedback mutex chain to the corresponding data asset objects respectively, record the starting asset object, ending asset object, mutex type, associated task constraint elements, and propagation direction of the mutex propagation subchain, and write the binding results into the candidate asset group.

[0063] A cross-asset mutual exclusion transmission chain is generated based on the mapped mutual exclusion transmission subchain, specifically as follows: Based on the mutual exclusion transmission sub-chain connection relationship between data asset objects within the candidate asset group, data asset objects with common task constraint elements, common output objects, or common calling paths are linked together to generate a cross-asset mutual exclusion transmission chain. The link nodes, link directions, mutual exclusion types, and associated same-asset evidence reversal factors are recorded in the cross-asset mutual exclusion transmission chain.

[0064] In this implementation scheme, the impact of the evidence reversal result of a data asset object on the calling order, replacement relationship and blocking relationship of other data asset objects in the candidate asset group is described by the cross-asset mutual exclusion transmission chain. This enables the blocking evidence within a single asset to be transmitted to the cross-asset combined calling process, reducing the path deviation caused by selecting based solely on the quality status of a single asset during combined calling.

[0065] Specifically, the steps for generating the output path reconstruction package are as follows: The application output task constraint package, the same-asset evidence reversal factor, and the cross-asset mutual exclusion transmission chain are read as follows: Read the corresponding application output task constraint package based on the application output task identifier, read the corresponding same-asset evidence inversion factor based on the data asset object identifier, and read the cross-asset mutual exclusion transmission chain associated with the application output task.

[0066] Based on the same-asset evidence reversal factor and the cross-asset mutually exclusive transmission chain, the path type identifier is generated as follows: Based on the inversion type, supporting path, blocking path in the same asset evidence inversion factor and the mutual exclusion type, link direction, and associated task constraint elements in the cross-asset mutual exclusion transmission chain, determine the path processing type of the data asset object in the application output task and generate a path type identifier.

[0067] Data asset objects are categorized by call type based on path type identifiers to obtain call asset groups. These call asset groups characterize the call role of data asset objects within the application output path, specifically: The data asset objects corresponding to the main call with the path type identifier are assigned to the main call asset group, the data asset objects that need to be verified are assigned to the delayed verification asset group, the data asset objects with blocking relationships are assigned to the abnormal blocking asset group, and the data asset objects that can replace the blocked data asset objects are assigned to the replacement call asset group. In one implementation, when a data asset object forms an abnormal blocked path due to permission boundaries, and there is another data asset object in the candidate asset group with the same field scope and that meets the permission boundary requirements, the other data asset object is assigned to the replacement call asset group.

[0068] The main call path, delayed verification path, exception blocking path, and replacement call path are generated according to the asset group being called, as follows: Generate the main call path based on the main call asset group, generate the delayed verification path based on the delayed verification asset group, generate the abnormal blocking path based on the abnormal blocking asset group, and generate the alternative call path based on the alternative call asset group. Then bind each path with the application output task identifier, the call asset group identifier, and the task constraint element identifier.

[0069] The application output path refactoring package is generated based on the main call path, delayed verification path, exception blocking path, and replacement call path, specifically as follows: The main call path, delayed verification path, exception blocking path, and alternative call path are arranged according to the execution order of the application output tasks to form an application output path refactoring package. The application output path refactoring package records the path order, path type, call asset group, task constraint elements, call interface, and output object.

[0070] Write the path type identifier into the application output path refactoring package, specifically as follows: Associate the path type identifier with the corresponding data asset object identifier, calling asset group identifier, and path identifier, and write it into the application output path reconstruction package so that the application output path reconstruction package can record the calling role and processing method of each data asset object in the application output task.

[0071] In this implementation plan, the same-asset evidence reversal factor and cross-asset mutual exclusion transmission chain are transformed into main call, delayed verification, abnormal blocking and replacement call path arrangements by applying the output path reconstruction package. This enables the application output task to reorganize the call path according to the support status and blocking status of governance evidence, reducing the problem that governance results remain in the registration and retrieval stage and are difficult to participate in output execution.

[0072] Specifically, the processing steps for the responsibility-based write-back update module are as follows: Collect path feedback data corresponding to the application output path reconstruction package. The path feedback data includes application output results, call feedback, and exception records, specifically: After the application output path reconstruction package is executed, collect whether the output is completed, whether the output fields are generated, whether the API call returns, whether the asset group call is successfully executed, whether the abnormal blocking path is triggered, whether the replacement call path is enabled, and user feedback records. Compile the above information into path feedback data.

[0073] Based on the application output path, the reconstructed package locates the data asset object corresponding to the path feedback data, which is specifically: Read the path identifier, called asset group identifier, task constraint element identifier, and exception record identifier from the path feedback data, locate the corresponding data asset object in the application output path reconstruction package, and establish a feedback mapping relationship between the path feedback data and the data asset object.

[0074] Based on the data asset object, the corresponding global data asset governance status package, the same-asset evidence reversal factor, and the cross-asset mutual exclusion transmission chain are read, specifically as follows: Based on the data asset object identifier in the feedback mapping relationship, read the corresponding global data asset governance status package, and simultaneously read the same asset evidence reversal factor and cross-asset mutual exclusion transmission chain associated with the data asset object.

[0075] The path feedback data is written into the supporting and limiting evidence surfaces of the global data asset governance status package, specifically as follows: When the path feedback data indicates that the data asset object has completed the call and met the task constraint elements, the corresponding feedback will be written into the supporting evidence surface. When the path feedback data indicates that the data asset object has experienced call failure, missing output, permission blocking, time expiration, quality abnormality, or replacement call, the corresponding feedback will be written into the restricted evidence surface. In one implementation, when the replacement call path is enabled and the replacement call asset group completes its output, the success feedback of the replacement call is written to the supporting evidence surface of the replacement data asset object, and the blocking feedback of the original replaced data asset object is written to the limiting evidence surface.

[0076] The asset reversal factor is updated based on the supporting and limiting evidence surfaces after writing, specifically as follows: Read the updated supporting and limiting evidence surfaces, determine whether there are supporting and blocking paths for the same data asset object that simultaneously point to the same application output task, and update the evidence reversal position, supporting path identifier, blocking path identifier, and reversal type in the same asset evidence reversal factor when the evidence reversal position changes.

[0077] The updated cross-asset mutual exclusion transmission chain and application output path reconstruction package are based on the updated same-asset evidence reversal factor, specifically as follows: Match the mutually exclusive propagation subchains in the candidate asset group according to the updated same-asset evidence reversal factor, and update the cross-asset mutually exclusive propagation chain when the mutually exclusive propagation subchain changes. Generate path type identifiers, calling asset groups, and calling paths based on the updated cross-asset mutual exclusion propagation chain, and update the application output path refactoring package.

[0078] In this implementation plan, the application output results are written back to the supporting evidence surface and the limiting evidence surface through path feedback data. This allows the feedback after the application output task is executed to continue to participate in the update of the same asset evidence reversal factor, cross-asset mutual exclusion transmission chain and application output path reconstruction package, so that data asset governance, application output and feedback write-back form a continuous processing link.

[0079] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0080] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A management system for the integrated governance and precise application output of data assets across the entire domain, characterized in that, include: The governance status package generation module is used to extract governance status elements and evidence data based on the full-domain data asset records, and generate a full-domain data asset governance status package. The task constraint package generation module is used to extract basic task elements and task constraint elements based on the application output task, and generate the application output task constraint package. The dual-state evidence splitting module is used to split the global data asset governance status package into supporting evidence and restrictive evidence based on the global data asset governance status package and the application output task constraint package. The constraint mirror graph generation module is used to construct the original governance dimension relationship graph based on the supporting evidence surface, the limiting evidence surface and the application output task constraint package, and to combine and reorganize the cross-dimensional node states into a constraint mirror graph. The evidence reversal identification module is used to identify the location in the constraint mirror diagram where the supporting evidence surface of the same data asset is blocked by the restricted evidence surface in reverse, and to generate the same asset evidence reversal factor. The conflict evidence transmission module is used to perform cross-asset transmission based on the same asset evidence reversal factor in the constraint mirror diagram, and generate cross-asset mutually exclusive transmission chains. The output path reconstruction module is used to reconstruct the data asset call path corresponding to the application output task based on the same asset evidence reversal factor and cross-asset mutual exclusion transmission chain, and generate the application output path reconstruction package. The Responsibility Write-back Update Module is used to update the global data asset governance status package, the same-asset evidence reversal factor, the cross-asset mutual exclusion transmission chain, and the application output path reconstruction package based on the application output results, call feedback, and exception records corresponding to the application output path reconstruction package.

2. The management system for the integrated governance and precise application output of data assets across the entire domain, as described in claim 1, is characterized in that... The steps for generating the global data asset governance status package are as follows: Read the full-domain data asset records, which include business domain data records, device-side data records, interface-side data records, document-side data records, indicator-side data records, and collaborative access data records; Merge asset objects from all data asset records across the domain to generate data asset objects; Extract the governance status elements of data asset objects. The governance status elements include source ownership, field definition, update cycle, quality rules, permission boundaries, and call records. Perform computer vision recognition on document-side data records to generate evidence data, which includes page layout evidence, table area evidence, and image field evidence. The governance status elements and evidence data are encapsulated into a global data asset governance status package.

3. The management system for the integrated governance and precise application output of data assets across the entire domain, as described in claim 2, is characterized in that... The steps of computer vision recognition are as follows: Read image files from document-side data records; Perform page layout segmentation on the image file to obtain page layout regions; Extract table field information and image field information from the layout area; Extract evidence of field structure from table field information; Extract image object evidence from image field information; Write the field structure evidence and image object evidence into the evidence data.

4. The management system for the integrated governance and precise application output of data assets across the entire domain, as described in claim 1, is characterized in that... The steps for separating supporting and limiting evidence are as follows: Read the governance status elements and evidence data from the global data asset governance status package; Read the task constraint elements from the application's output task constraint package; Write the governance status elements and evidence data that meet the task constraints into the supporting evidence surface; Write the governance status elements and evidence data of the blocking task constraint elements into the restricted evidence surface; The supporting and limiting evidence are bound to the same data asset object, and a bi-state evidence index is generated based on the binding result.

5. The management system for the integrated governance and precise application output of data assets across the entire domain, as described in claim 1, is characterized in that... The steps for generating the application output task constraint package are as follows: Analyze the application output task to obtain the basic elements of the task, which include the application scenario, the output object, and the output format. Task constraint elements are generated based on basic task elements. These constraints include indicator definitions, field ranges, permission boundary requirements, invocation constraints, timeliness requirements, and quality requirements. The basic elements and constraints of the task are encapsulated into an application output task constraint package.

6. The management system for the integrated governance and precise application output of full-domain data assets according to claim 1, characterized in that, The steps for generating a constraint mirror diagram are as follows: The supporting and limiting evidence will be established as dimensional nodes based on governance status elements and evidence data. The application output task constraint package is used to create task nodes according to the basic task elements and task constraint elements. Establish the original governance dimension relationship diagram between dimension nodes and task nodes; Extract the state combinations of cross-dimensional connected nodes from the original governance dimension relationship graph to generate mirror nodes; Mirror edges are generated based on the support and blocking relationships between mirror nodes, and a constrained mirror graph is generated based on the mirror nodes and mirror edges.

7. The management system for the integrated governance and precise application output of data assets across the entire domain, as described in claim 1, is characterized in that... The steps for generating the asset-backed evidence reversal factor are as follows: Read the supporting evidence surface, limiting evidence surface, and application output task constraint package corresponding to the same data asset object in the constraint mirror diagram; Find the supporting evidence surface along the mirror edge to match the supporting path of the application output task constraint package; Find the blocking path in the application output task constraint package by searching along the mirror edge to restrict the evidence surface; When both the supporting path and the blocking path point to the same application output task, mark the evidence reversal position; Generate an evidence reversal factor for the same asset based on the evidence reversal location, supporting path, and blocking path.

8. The management system for the integrated governance and precise application output of full-domain data assets according to claim 1, characterized in that, The steps for generating a cross-asset mutual exclusion transmission chain are as follows: Read the data asset objects that match the task constraint elements in the constraint mirror diagram and generate candidate asset groups; Read the data asset object corresponding to the asset evidence reversal factor; Extract governance status elements from candidate asset groups; Based on the overlapping states between governance status elements, mutually exclusive propagation subchains are generated. Mutually exclusive propagation subchains are used to represent the constraint mutual exclusion propagation relationship between different data asset objects. Mutually exclusive propagation subchains include time-limited mutual exclusion chains, call mutual exclusion chains, and feedback mutual exclusion chains. Map the mutually exclusive propagation subchain to the data asset objects in the candidate asset group; Generate a cross-asset mutual exclusion transmission chain based on the mapped mutual exclusion transmission subchain.

9. The management system for the integrated governance and precise application output of full-domain data assets according to claim 1, characterized in that, The steps for generating the application output path reconstruction package are as follows: Read the application output task constraint package, same-asset evidence inversion factor, and cross-asset mutual exclusion transmission chain; Generate path type identifiers based on the same asset evidence reversal factor and cross-asset mutually exclusive transmission chains; Data asset objects are classified into call asset groups based on path type identifiers. Call asset groups are used to represent the call role of data asset objects in the application output path. Generate the main call path, delayed verification path, exception blocking path, and alternative call path according to the asset group being called; Generate an application output path reconstruction package based on the main call path, delayed verification path, exception blocking path, and replacement call path; Write the path type identifier into the application output path reconstruction package.

10. The management system for the integrated governance and precise application output of full-domain data assets according to claim 1, characterized in that, The processing steps for the responsibility-based write-back update module are as follows: Collect path feedback data corresponding to the application output path reconstruction package. The path feedback data includes application output results, call feedback and exception records. Based on the application output path, reconstruct the package to locate the data asset object corresponding to the path feedback data; Based on the data asset object, read the corresponding global data asset governance status package, same-asset evidence reversal factor, and cross-asset mutual exclusion transmission chain; Write the path feedback data into the supporting and limiting evidence aspects of the global data asset governance status package; The asset evidence reversal factor is updated based on the supporting and limiting evidence surfaces after writing. The cross-asset mutual exclusion transmission chain and application output path reconstruction package are updated based on the updated same-asset evidence inversion factor.

Images