Penetrating work and full-level control generation method based on multi-source heterogeneous data

Abstract

The application relates to the technical field of data processing, and particularly discloses a penetrating work and full-level control generation method based on multi-source heterogeneous data, which comprises the following steps: taking data entities and entity attribute information in original data from multiple heterogeneous data sources as nodes, and taking the association relationship of semantic similarity between the data entities as edges to construct a multi-source heterogeneous association graph of a commission-enterprise informationization platform; performing hierarchical penetrating path search on the multi-source heterogeneous association graph, and determining a penetrating control strategy set of the commission-enterprise informationization platform; according to a penetrating work request initiated by a user, sequentially accessing nodes at all levels on a work path, calling corresponding control strategies in the penetrating control strategy set to perform compliance verification when each node is accessed, and collecting execution logs of all the nodes to obtain a full-level control result report of the commission-enterprise informationization platform; and the application can improve the efficiency of the penetrating work and the full-level control.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method for generating comprehensive management and control based on multi-source heterogeneous data. Background Technology

[0002] In the existing information management and control scenarios of government agencies and enterprises, there is a lack of unified association modeling standards for multi-source heterogeneous data. Data entities and attributes cannot be automatically identified and integrated to achieve semantic association, making it difficult to build a multi-source heterogeneous association map that connects the entire domain. In addition, the data standards of government agencies and enterprises are not unified and are stored in a scattered manner. Regulatory data relies on periodic manual reporting, resulting in problems such as data delay and lack of consistency. This makes it impossible to provide real-time and complete data support for penetrating work, and the efficiency of data integration and association analysis is low.

[0003] Traditional penetrating control and full-level supervision lack an automated hierarchical penetration path search mechanism. Control strategies cannot be accurately matched according to organizational level and business type. Compliance verification relies on manual operation, which is cumbersome and slow. The efficiency of execution log summary and control report generation is low. At the same time, the dual full coverage of supervision business and supervised enterprises is slow. Intelligent supervision lacks multi-dimensional data empowerment. The cost of repeated investment in information technology construction is high. The real-time performance, accuracy and automation level of the overall control are difficult to meet the needs of the commission and enterprises at all levels. Summary of the Invention

[0004] This invention provides a method for generating a comprehensive working and full-level control system based on multi-source heterogeneous data to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for penetrating and managing multi-level control based on multi-source heterogeneous data, comprising:

[0006] Step 1: Using data entities and entity attribute information from raw data sources from multiple heterogeneous data sources as nodes and semantic similarity relationships between data entities as edges, construct a multi-source heterogeneous association graph for the enterprise information platform.

[0007] Step 2: Based on the query constraints of the penetration query request, perform a hierarchical penetration path search on the multi-source heterogeneous association graph, and determine the set of penetration control strategies for the enterprise information platform based on the node level and business type of each searched work path.

[0008] Step 3: Based on the user-initiated penetrating work request, access each level of nodes on the work path in sequence. When accessing each node, call the corresponding control policy in the penetrating control policy set for compliance verification. Summarize the execution logs of all nodes to obtain the full-level control result report of the enterprise information platform.

[0009] Preferably, the construction of a multi-source heterogeneous association graph for the enterprise information platform, using data entities and entity attribute information from multiple heterogeneous data sources as nodes and semantic similarity relationships between data entities as edges, includes:

[0010] The raw data extracted from the heterogeneous data source is parsed and identified to obtain the data entities and corresponding entity attribute information in the raw data, and a temporary entity identifier is generated for each data entity.

[0011] Based on temporary entity identifiers, calculate the semantic similarity between data entities and their corresponding entity attribute information;

[0012] When the semantic similarity exceeds a preset threshold, it is determined that there is a relationship between the two corresponding data entities;

[0013] Using data entities as nodes and relationships as edges, a multi-source heterogeneous relationship graph is constructed for the enterprise information platform.

[0014] Preferably, the step of parsing and identifying the raw data extracted from the heterogeneous data source includes:

[0015] Obtain the full-level organizational structure list of the enterprise information platform. The full-level organizational structure list includes the names of all nodes from the highest management level to the grassroots operating units and the reporting relationships between nodes.

[0016] The business department identifier and business level identifier of the original data recorded in each heterogeneous data source will be read and bound to the corresponding organizational node to form an organizational node mapping table for the heterogeneous data source.

[0017] Based on the organization node mapping table, query the data source type and data source identifier of the entity in each heterogeneous data source to obtain the node path and associated entities of the organization node. The associated entities include regulatory object entities, business matter entities, regulatory indicator entities and regulatory rule entities.

[0018] When the same organization node is associated with entities from multiple heterogeneous data sources, all entities belonging to the organization node are merged into a comprehensive entity set of the organization node;

[0019] The original source tags of each source entity are retained within the comprehensive entity set to distinguish the data differences in the description of the same organizational node by different business platforms.

[0020] Preferably, the formula for calculating the semantic similarity is as follows:

[0021] ;

[0022] In the formula, To indicate the first The data entity and the first Semantic similarity between data entities For data entities, This represents the total number of key attribute types in the entity attribute information. For the first The preset weight coefficients of various attributes in semantic similarity calculation It is a cosine function. For the first The first data entity The first attribute vector obtained after vectorizing the attribute values ​​is the first attribute value. One portion, For the first The first data entity The second attribute vector obtained after the attribute value is vectorized One portion, For the first The set of contextual identifiers for each data entity appearing in the enterprise information platform. For the first The set of contextual identifiers for each data entity appearing in the enterprise information platform.

[0023] Preferably, the step of performing hierarchical penetration path search on the multi-source heterogeneous association graph based on the query constraints of the penetration query request includes:

[0024] Accepting penetrating query requests initiated by users through the enterprise-commission linkage system, the parsed target organization node identifier, penetration depth level, and the type of business matter to be queried are aggregated into query constraints for a multi-source heterogeneous association graph;

[0025] Based on query constraints, starting from the target organization node identifier, the corresponding starting node is located in the multi-source heterogeneous association graph, and the node attribute information of the starting node in the graph is obtained.

[0026] Based on the penetration depth level, starting from the starting node, perform a breadth-first traversal along the reporting membership edges in the graph to the next lower level node.

[0027] Preferably, the breadth-first traversal includes:

[0028] When traversing each level, the nodes visited at the current level are recorded as path nodes, until the deepest level specified by the penetration depth level number is reached or the current level has no lower-level nodes, then the traversal stops.

[0029] During the traversal, based on the type of business matter to be queried, nodes in the entity set associated with each path node that contain entity information matching the type of business matter are retained as retained nodes.

[0030] Combine the reporting hierarchy edges between nodes into one or more complete hierarchical penetration paths.

[0031] Preferably, the step of determining the set of penetrating control strategies for the enterprise information platform based on the node level and business type of each work path searched includes:

[0032] Based on the organizational hierarchy attribute of the hierarchical penetration path, each node is matched with the permission control template corresponding to the current level. The permission control template defines the data viewing scope, data modification permissions, and data reporting requirements allowed for the corresponding level node.

[0033] Based on the type of business matter, a corresponding rule control template is matched for each node. The rule control template includes threshold comparison rules, data integrity verification rules, and reporting timeliness verification rules for business matters under the commission-enterprise linkage system.

[0034] The permission control template and the rule control template will be used together as the single-node control strategy in this penetration query.

[0035] Use the hierarchical penetration path as the working path;

[0036] The control strategies for each node are statistically summarized according to the order of nodes in the work path, resulting in a set of penetrating control strategies for the enterprise information platform.

[0037] Preferably, the step of sequentially accessing nodes at each level of the work path according to the user-initiated penetrating work request, and calling the corresponding control policy in the penetrating control policy set for compliance verification when accessing each node, includes:

[0038] Extract the target work path identifier and the type of operation to be performed from the user-initiated through-work request, and determine that the target work path identifier points to an already generated work path;

[0039] According to the order of the corresponding nodes in the work path, the process starts from the first node and proceeds downwards. When a node is reached, the subsequent access is paused and the process switches to the strategy invocation phase.

[0040] Based on the node identifier of the corresponding node and the type of operation to be executed, retrieve the matching control policy from the set of penetrating control policies;

[0041] Execute the retrieved control policy, complete the compliance verification of the corresponding node, and decide whether to continue accessing the next node on the path or terminate the entire work request based on the verification results and the feedback data after execution.

[0042] Preferably, the compliance verification of the current node includes:

[0043] Read the data access permission rules stored in the metadata of the corresponding node. The data access permission rules include a list of user roles that are allowed to access the corresponding node and a list of operations that are allowed to be performed.

[0044] Based on the user's role in the penetrating work request, the user role is matched with the list of user roles. If the match fails, the verification is directly determined to be unsuccessful and subsequent access is denied.

[0045] After a successful role match, further check whether the type of operation to be performed is included in the list of allowed operations. If it is not included, the verification fails and the operation of the current node is terminated.

[0046] If both the role and the operation type match successfully, a compliance record containing the current node identifier, user identifier, operation timestamp, and operation type will be generated according to the operation record requirements of the penetration-type work request. The compliance record and the post-execution feedback data will be temporarily stored in the verification result cache.

[0047] After all corresponding nodes have been accessed, all compliance records in the cache and post-execution feedback data are aggregated into the execution log of the penetrating work request.

[0048] Preferably, the summary of execution logs from all nodes to obtain a comprehensive control result report for the enterprise information platform includes:

[0049] According to the hierarchical order of the nodes in the working path, sort and merge the execution logs corresponding to all nodes to obtain the log sequence of the nodes;

[0050] Based on the verification result status and post-execution feedback data of each execution log in the log sequence, count the number of nodes that passed verification and the number of nodes that failed verification on the working path, and mark the specific level position of the nodes that failed verification.

[0051] The log sequence, statistical results, and marked non-passing node information are assembled into a template to generate a full-level control result report for the enterprise information platform.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. This method constructs a multi-source heterogeneous association graph to achieve unified association modeling of data entities and attributes, automatically completes the semantic association identification and integration between entities, opens up the data linkage channel between the commission and enterprises, improves the efficiency of data integration, retrieval and association analysis, provides stable and real-time data support for penetrating work, and ensures the consistency and accuracy of regulatory data.

[0054] 2. This method relies on an automated hierarchical penetration path search and precise control strategy matching mechanism to simplify the compliance verification process, improve the efficiency of execution log aggregation and full-level control report generation, realize the automated operation of penetration work and full-level control, comprehensively improve the timeliness, accuracy and intelligence of control, and support the efficient implementation of the commission-enterprise linkage full-domain online supervision. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating a method for generating a comprehensive, multi-level control system based on multi-source heterogeneous data, according to an embodiment of the present invention.

[0056] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0057] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0058] This application provides a method for generating a comprehensive workflow and full-level control based on multi-source heterogeneous data. The execution entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0059] Reference Figure 1 The diagram shown is a flowchart illustrating a method for generating comprehensive working processes and full-level control based on multi-source heterogeneous data, according to an embodiment of the present invention. In this embodiment, the method for generating comprehensive working processes and full-level control based on multi-source heterogeneous data includes:

[0060] Step 1: Using data entities and entity attribute information from raw data sources from multiple heterogeneous data sources as nodes and semantic similarity relationships between data entities as edges, construct a multi-source heterogeneous association graph for the enterprise information platform.

[0061] In this embodiment of the invention, the construction of a multi-source heterogeneous association graph for the enterprise information platform, using data entities and entity attribute information from multiple heterogeneous data sources as nodes and semantic similarity relationships between data entities as edges, includes:

[0062] The raw data extracted from the heterogeneous data source is parsed and identified to obtain the data entities and corresponding entity attribute information in the raw data, and a temporary entity identifier is generated for each data entity.

[0063] Based on temporary entity identifiers, calculate the semantic similarity between data entities and their corresponding entity attribute information;

[0064] When the semantic similarity exceeds a preset threshold, it is determined that there is a relationship between the two corresponding data entities;

[0065] Using data entities as nodes and relationships as edges, a multi-source heterogeneous relationship graph is constructed for the enterprise information platform.

[0066] The step of parsing and identifying the raw data extracted from heterogeneous data sources includes:

[0067] Obtain the full-level organizational structure list of the enterprise information platform. The full-level organizational structure list includes the names of all nodes from the highest management level to the grassroots operating units and the reporting relationships between nodes.

[0068] The business department identifier and business level identifier of the original data recorded in each heterogeneous data source will be read and bound to the corresponding organizational node to form an organizational node mapping table for the heterogeneous data source.

[0069] Based on the organization node mapping table, query the data source type and data source identifier of the entity in each heterogeneous data source to obtain the node path and associated entities of the organization node. The associated entities include regulatory object entities, business matter entities, regulatory indicator entities and regulatory rule entities.

[0070] When the same organization node is associated with entities from multiple heterogeneous data sources, all entities belonging to the organization node are merged into a comprehensive entity set of the organization node;

[0071] The original source tags of each source entity are retained within the comprehensive entity set to distinguish the data differences in the description of the same organizational node by different business platforms.

[0072] The formula for calculating semantic similarity is as follows:

[0073] ;

[0074] In the formula, To indicate the first The data entity and the first Semantic similarity between data entities For data entities, This represents the total number of key attribute types in the entity attribute information. For the first The preset weight coefficients of various attributes in semantic similarity calculation It is a cosine function. For the first The first data entity The first attribute vector obtained after vectorizing the attribute values ​​is the first attribute value. One portion, For the first The first data entity The second attribute vector obtained after the attribute value is vectorized One portion, For the first The set of contextual identifiers for each data entity appearing in the enterprise information platform. For the first The set of contextual identifiers for each data entity appearing in the enterprise information platform.

[0075] The system reads raw data from heterogeneous data sources one by one. For each raw data, it performs structural decomposition and content parsing according to the data storage format. It identifies objects in the data that can independently carry business meaning as data entities. Simultaneously, it extracts the descriptive, attribution, and constraint information corresponding to each data entity as entity attribute information. According to the extraction order of data entities and combined with the data source encoding, it generates a fixed-length unique character code for each data entity as a temporary entity identifier, ensuring that each data entity corresponds to only one non-repeating temporary entity identifier.

[0076] Using temporary entity identifiers as the matching basis, the data entity corresponding to each temporary entity identifier and all entity attribute information bound to the data entity are accurately located. The core business text description content of the data entity is extracted, and then the business text description content of the entity attributes is extracted item by item. The matching of the two in terms of business orientation, organizational scope, and business application scenario is compared one by one. The number of key description items that are completely matched is counted. The ratio obtained by dividing the number of key description items that are completely matched by the total number of key description items is used as the semantic similarity between the data entity and the corresponding entity attribute information.

[0077] The semantic similarity threshold is pre-set to 0.8. The calculated semantic similarity value is then compared with 0.8. When the calculated semantic similarity value is greater than 0.8, it is directly determined that the two data entities corresponding to that semantic similarity value have a relationship.

[0078] All data entities that have been parsed and identified are uniformly used as nodes in the multi-source heterogeneous association graph of the enterprise information platform. The correspondence between all two data entities that are determined to have a relationship is used as the edge of the multi-source heterogeneous association graph of the enterprise information platform. The nodes and edges are arranged and connected according to the business association logic and organizational affiliation logic, and finally the multi-source heterogeneous association graph of the enterprise information platform is fully constructed.

[0079] The system configuration database of the Jilin Provincial State-owned Assets Supervision and Administration Commission (SASAC) and State-owned Enterprises Information Platform was used to retrieve full-level organizational data of provincial and municipal SASACs and their affiliated state-owned enterprises. The data was then systematically analyzed according to the management hierarchy: Provincial SASAC, Municipal SASACs, Group Enterprises, Subsidiaries, and Grassroots Operational Units. A complete list of all organizational nodes from the highest management level (Provincial SASAC) to the grassroots operational units was compiled. The reporting relationships between each organizational node and its direct superiors and subordinates were recorded, forming a complete list of the full-level organizational structure of the SASAC and State-owned Enterprises Information Platform. Data was read from heterogeneous data sources such as the SASAC Online Supervision Platform, the SASAC Big Data Center, and the State-owned Enterprises Business Information Platform. The business department identifier and business level identifier fields were precisely extracted from each piece of raw data. These extracted business department identifiers and business level identifiers were then bound one-to-one with the matching organizational nodes in the full-level organizational structure list. A structured organizational node mapping table for the heterogeneous data sources was then created based on the correspondence between data source number, business department identifier, business level identifier, and organizational node name.

[0080] Based on the generated organizational node mapping table of heterogeneous data sources, all entities contained in heterogeneous data sources such as the State-owned Assets and Enterprises Online Supervision Platform and the State-owned Enterprises Business Information Platform are located one by one. The data source type and data source identifier corresponding to each entity are traced and determined. The complete node path of the organizational node to which the entity belongs is sorted out through the hierarchical relationship of the organizational nodes. At the same time, the entities are divided into four categories: regulatory object entities, business matter entities, regulatory indicator entities, and regulatory rule entities. Finally, the node path and associated entities corresponding to each organizational node are obtained.

[0081] The number of data source sources associated with each organizational node is checked one by one. When it is determined that the entities associated with a certain organizational node come from two or more different heterogeneous data sources, such as the State-owned Assets and Enterprises Online Supervision Platform, the State-owned Enterprises Business Information Platform, and the State-owned Assets and Enterprises Big Data Center, all heterogeneous data source entities corresponding to that organizational node are centrally collected and integrated to form a comprehensive entity set exclusive to that organizational node.

[0082] After merging and aggregating the comprehensive entity set, retain the original data source tag for each entity in the comprehensive entity set. This tag is used to mark the state-owned assets supervision cloud platform, provincial and municipal state-owned assets supervision platforms, or state-owned enterprise business information platforms to which the entity originally belonged. The tag distinguishes the differences in data content and descriptions of the same organizational node described by different business platforms.

[0083] The total number of key attribute types is obtained by screening the core attributes of the entity attribute information corresponding to each data entity. The screening criteria are strictly based on the core dimensions of the state-owned asset supervision business of the enterprise information platform, covering core judgment dimensions such as organizational affiliation, business type, regulatory level, and data status. Finally, the total number of attribute types that meet the requirements is counted to determine this parameter.

[0084] The preset weighting coefficients are calculated by combining the importance level of online supervision of state-owned assets and enterprises, the influence of each attribute in historical supervision data on entity association determination, and the priority of supervision business at all levels of the commission-enterprise linkage. Fixed values ​​were determined after repeated verification using entity association matching results from historical supervision cases, with core supervision attributes corresponding to higher weighting coefficients. Attribute vector components are obtained by converting the attribute values ​​of data entities into standardized numerical vectors according to the unified data standardization specifications of the commission-enterprise information platform. The numerical components corresponding to each single attribute are then extracted from the vectors. All attribute values ​​from heterogeneous data sources use the same conversion rule to ensure complete consistency in the numerical definitions of the components.

[0085] The context identifier set extracts the identification information of data entities within the enterprise information platform, including their business module, organizational node, regulatory scenario, and data source type. This identification information is then integrated into a dedicated set. All identification information originates directly from the mapping table between original data tags from heterogeneous data sources and organizational nodes. Data entities, obtained through parsing and identification of multi-source heterogeneous data sources, are the core objects carrying business meaning within the enterprise information platform. Specifically, they include four categories: regulatory object entities, business matter entities, regulatory indicator entities, and regulatory rule entities. All are business carriers extracted after binding to the full-level organizational structure.

[0086] This formula is used to quantify the semantic similarity between two data entities, providing an accurate basis for determining associations in the construction of multi-source heterogeneous association graphs for the enterprise information platform. By integrating the weighted matching results of attributes with the overlap results of the context environment, it measures the degree of entity association in line with the business logic of penetrating supervision of state-owned assets and enterprises, providing stable data association support for subsequent hierarchical penetrating path search and penetrating control strategy matching.

[0087] The beneficial effects include the ability to accurately analyze heterogeneous data sources in the state-owned assets and enterprises sector, generate unique temporary identifiers for data entities, accurately calculate semantic similarity between entities and determine their relationships, efficiently build multi-source heterogeneous association graphs, complete the binding mapping between organizational nodes and data sources based on a full-level organizational structure list, integrate entities from multiple sources to form a comprehensive entity set while retaining the original source labels, and open up data linkage channels between the commission and enterprises. All parameters in the formula are determined based on the dimensions of state-owned assets supervision business and historical data, which can accurately quantify the semantic association degree of data entities, provide a reliable basis for association graph construction, and achieve unified association modeling of multi-source heterogeneous data. This significantly improves the efficiency of data integration and association analysis, provides stable data support for hierarchical penetration path search, control strategy matching, and full-level penetration supervision, and fully adapts to the full-domain online supervision needs of state-owned assets and enterprises and commission-enterprise linkage.

[0088] Step 2: Based on the query constraints of the penetration query request, perform a hierarchical penetration path search on the multi-source heterogeneous association graph, and determine the set of penetration control strategies for the enterprise information platform based on the node level and business type of each searched work path.

[0089] In this embodiment of the invention, the step of performing a hierarchical penetration path search on the multi-source heterogeneous association graph based on the query constraints of the penetration query request includes:

[0090] Accepting penetrating query requests initiated by users through the enterprise-commission linkage system, the parsed target organization node identifier, penetration depth level, and the type of business matter to be queried are aggregated into query constraints for a multi-source heterogeneous association graph;

[0091] Based on query constraints, starting from the target organization node identifier, the corresponding starting node is located in the multi-source heterogeneous association graph, and the node attribute information of the starting node in the graph is obtained.

[0092] Based on the penetration depth level, starting from the starting node, perform a breadth-first traversal along the reporting membership edges in the graph to the next lower level node.

[0093] The breadth-first traversal includes:

[0094] When traversing each level, the nodes visited at the current level are recorded as path nodes, until the deepest level specified by the penetration depth level number is reached or the current level has no lower-level nodes, then the traversal stops.

[0095] During the traversal, based on the type of business matter to be queried, nodes in the entity set associated with each path node that contain entity information matching the type of business matter are retained as retained nodes.

[0096] Combine the reporting hierarchy edges between nodes into one or more complete hierarchical penetration paths.

[0097] The process involves determining a set of penetrating control strategies for the enterprise information platform based on the node level and business type of each work path found through the search, including:

[0098] Based on the organizational hierarchy attribute of the hierarchical penetration path, each node is matched with the permission control template corresponding to the current level. The permission control template defines the data viewing scope, data modification permissions, and data reporting requirements allowed for the corresponding level node.

[0099] Based on the type of business matter, a corresponding rule control template is matched for each node. The rule control template includes threshold comparison rules, data integrity verification rules, and reporting timeliness verification rules for business matters under the commission-enterprise linkage system.

[0100] The permission control template and the rule control template will be used together as the single-node control strategy in this penetration query.

[0101] Use the hierarchical penetration path as the working path;

[0102] The control strategies for each node are statistically summarized according to the order of nodes in the work path, resulting in a set of penetrating control strategies for the enterprise information platform.

[0103] The system receives penetrating query requests initiated by users within the State-owned Assets Supervision and Administration Commission (SASAC) and State-owned Enterprises (SOE) linkage system. It then breaks down and parses the content of the query request field by field, extracting the target organizational node identifier used to locate the regulatory organizational unit, the number of penetration depth levels used to limit the scope of regulatory viewing, and the type of business matter to be queried used to define the scope of regulatory business. The system integrates and summarizes the above three parsing results to form the query constraints of the multi-source heterogeneous association graph of the SOE information platform.

[0104] Based on the established query constraints, the target organization node identifier is used as the retrieval basis. The identifier information of each node is matched and compared one by one in the multi-source heterogeneous association graph of the enterprise information platform to accurately find the starting node that is completely consistent with the target organization node identifier. At the same time, all node attribute information stored in the graph, such as organizational level, business type, related entities, and affiliation, of the starting node are retrieved.

[0105] According to the number of penetration depth levels specified in the query constraints, starting from the located starting node, relying on the reporting and affiliation edges already constructed in the multi-source heterogeneous association graph, the traversal operation is performed on all subordinate organizational nodes level by level by level, using the method of first completing the traversal of all nodes in the current level and then entering the next level, to complete the breadth-first traversal.

[0106] When traversing each organizational level of the multi-source heterogeneous association graph of state-owned assets and enterprises, all organizational nodes that have completed the access operation within that level are registered one by one and marked as path nodes. The hierarchical traversal operation is continuously executed until the traversal level reaches the deepest regulatory level limited by the number of penetration depth levels, or when there are no lower-level organizational nodes in the currently traversed organizational level, the entire traversal operation is immediately terminated.

[0107] Throughout the entire process of traversing the organizational hierarchy, the business item type to be queried is retrieved from the query constraints. The comprehensive entity set bound to each marked path node at each level is checked one by one. Only the path nodes containing entity information that is completely consistent with the business item type to be queried are determined as the retained nodes for this penetration query, and the other mismatched path nodes are directly excluded.

[0108] All selected nodes are connected in an orderly manner according to the established reporting affiliation edges in the multi-source heterogeneous association graph. The selected nodes are combined with the corresponding reporting affiliation edges to form one or more complete hierarchical penetration paths that meet the needs of penetration query.

[0109] The organization level attribute of each node in the hierarchical penetration path is read, and the permission control template that is completely matched with the organization level attribute is retrieved from the permission control template library built into the enterprise information platform. The template clearly defines the data viewing scope, data modification permissions and data reporting requirements that the corresponding level node can access.

[0110] Extract the business item type bound to each node in the hierarchical penetration path, and retrieve the rule control template that is completely matched to the business item type from the rule control template library built into the State-owned Assets Supervision and Administration Commission (SASAC) information platform. The template contains threshold comparison rules, data integrity verification rules, and reporting timeliness verification rules executed under the SASAC linkage system.

[0111] The permission control template and rule control template matched for a single node will be integrated, and the integrated complete content will serve as the single node control strategy for this penetrating query.

[0112] The hierarchical penetration path generated in the early stage through breadth-first traversal is directly determined as the working path corresponding to this penetration work, preserving the original arrangement order and reporting hierarchy of the nodes in the path. According to the order of the nodes in the working path, the single-node control strategy corresponding to each node is collected in sequence. All single-node control strategies are uniformly organized and summarized to finally generate the penetration control strategy set of the enterprise information platform.

[0113] The beneficial effects include the ability to accurately analyze penetrating query requests under the state-owned assets and state-owned enterprises (SOEs) linkage system and form standardized query constraints; quickly locate the starting node and complete breadth-first traversal according to rules; effectively filter and retain nodes that match business needs and generate complete hierarchical penetrating paths; accurately match permission control templates and rule control templates based on node organizational hierarchy and business type to form standardized single-node control strategies; and generate a complete set of penetrating control strategies after being systematically summarized according to work paths. This enables automatic search of hierarchical penetrating paths and accurate matching of control strategies, significantly improving the automation level and execution accuracy of penetrating supervision of SOEs and providing stable and reliable strategy support for compliance verification work at all levels.

[0114] Step 3: Based on the user-initiated penetrating work request, access each level of nodes on the work path in sequence. When accessing each node, call the corresponding control policy in the penetrating control policy set for compliance verification. Summarize the execution logs of all nodes to obtain the full-level control result report of the enterprise information platform.

[0115] In this embodiment of the invention, the step of sequentially accessing nodes at each level of the work path according to the user-initiated penetrating work request, and calling the corresponding control policy in the penetrating control policy set for compliance verification when accessing each node, includes:

[0116] Extract the target work path identifier and the type of operation to be performed from the user-initiated through-work request, and determine that the target work path identifier points to an already generated work path;

[0117] According to the order of the corresponding nodes in the work path, the process starts from the first node and proceeds downwards. When a node is reached, the subsequent access is paused and the process switches to the strategy invocation phase.

[0118] Based on the node identifier of the corresponding node and the type of operation to be executed, retrieve the matching control policy from the set of penetrating control policies;

[0119] Execute the retrieved control policy, complete the compliance verification of the corresponding node, and decide whether to continue accessing the next node on the path or terminate the entire work request based on the verification results and the feedback data after execution.

[0120] The aggregation of execution logs from all nodes yields a comprehensive control result report for the enterprise information platform, including:

[0121] According to the hierarchical order of the nodes in the working path, sort and merge the execution logs corresponding to all nodes to obtain the log sequence of the nodes;

[0122] Based on the verification result status and post-execution feedback data of each execution log in the log sequence, count the number of nodes that passed verification and the number of nodes that failed verification on the working path, and mark the specific level position of the nodes that failed verification.

[0123] The log sequence, statistical results, and marked non-passing node information are assembled into a template to generate a full-level control result report for the enterprise information platform.

[0124] The system receives penetrating work requests initiated by users within the State-owned Assets Supervision and Administration Commission (SASAC) and State-owned Enterprises (SOE) linkage system. It then parses and splits the request content field by field, accurately extracting the target work path identifier used to locate the execution path and the type of operation to be executed used to clarify the operation content. The extracted target work path identifier is then compared one by one with the identifiers of all generated work paths to confirm that the identifier uniquely corresponds to a work path that has been constructed.

[0125] Based on the established sequence of nodes within the confirmed work path, node access operation is initiated from the first node at the very beginning of the work path. The access process is then advanced step by step down the organizational hierarchy. Upon successfully reaching each node, access operations for subsequent nodes are immediately paused, and the process switches to the execution phase of the control policy invocation.

[0126] The system retrieves the unique node identifier of the currently accessed node and the type of operation to be executed in the penetrating work request. Using the node identifier and operation type as dual search criteria, it performs a one-by-one match search in the generated set of penetrating control strategies of the enterprise information platform to accurately locate the control strategy that is perfectly compatible with the current node and operation type.

[0127] Initiate the found adaptive control policy and fully execute the policy content. Complete the compliance verification operation of the current node according to the policy. Record the verification results and the feedback data generated after execution. If the verification result meets the preset compliance standard, continue to access the next node in the work path. If the verification result does not meet the preset compliance standard, immediately terminate the entire execution process of the penetration work request.

[0128] Retrieve the data access permission rules permanently stored in the metadata of the currently accessed node, and obtain a complete list of all user roles allowed to access this node and a complete list of all operations allowed to be performed by this node, ensuring that the permission rule information is completely extracted without any omissions.

[0129] Extract the fixed role of the user who initiated the penetration work request in the enterprise information platform, and compare the user role with the user role list in the node data access permission rules item by item. If the user role does not appear in the user role list, the compliance verification is directly determined to be unsuccessful, and all access operations of the user to the current node and subsequent nodes are denied.

[0130] After matching user roles with the user role list, the types of operations to be executed in the penetrating work request are checked item by item against the list of operations allowed to be executed by the node. If the type of operation to be executed is not included in the list of operations, the compliance check is deemed to have failed, and all operation execution processes of the current node are immediately terminated. After the user role is successfully matched and the type of operation to be executed is included in the list of operations allowed to be executed, a complete compliance record containing the current node identifier, user identifier, operation execution timestamp, and operation type is generated according to the operation record requirements set in the penetrating work request. This compliance record and the feedback data generated after the node executes the operation are synchronously stored in the verification result cache for temporary storage.

[0131] After completing the access and verification operations for all corresponding nodes on the work path, all compliance records and post-execution feedback data are extracted from the verification result cache, and integrated and collected in the order of node access to form a complete execution log for this penetrating work request.

[0132] Based on the predetermined hierarchical arrangement of organizational nodes in the work path from the upper management level to the lower work unit, the execution logs corresponding to all nodes covered by the work path are sequentially arranged and merged to form a continuous and complete node log sequence.

[0133] The system reads the verification result status and post-execution feedback data of each execution log record in the node log sequence one by one. It counts the number of nodes whose verification results meet the compliance standards among all nodes in the working path, and counts the number of nodes whose verification results do not meet the compliance standards. It accurately locates and marks the specific hierarchical position of each node that fails verification in the organizational structure of the enterprise information platform.

[0134] The completed node log sequence, the statistical results of the number of nodes that passed and failed verification, and the specific hierarchical location information of the nodes that failed verification are filled in and combined one by one according to the fixed format and filling position of the full-level control report template preset by the enterprise information platform, and finally a standardized full-level control result report of the enterprise information platform is generated.

[0135] The beneficial effects include the ability to accurately analyze penetrating work requests under the state-owned assets and state-owned enterprises (SOEs) linkage system, access nodes at all levels in an orderly manner according to the work path and match and adapt control strategies, strictly complete dual compliance verification of user roles and operation permissions, automatically generate compliance records and execution logs, orderly summarize logs to form a complete sequence and accurately statistically analyze verification results, and quickly generate standardized full-level control result reports. This significantly improves the automation level and accuracy of penetrating regulatory compliance verification, and ensures the efficient and standardized implementation of full-level control work of SOEs.

[0136] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0137] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0138] Finally, it should be noted that the above 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 preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for generating comprehensive management and control based on multi-source heterogeneous data, characterized in that: The method includes: Step 1: Using data entities and entity attribute information from raw data sources from multiple heterogeneous data sources as nodes and semantic similarity relationships between data entities as edges, construct a multi-source heterogeneous association graph for the enterprise information platform. Step 2: Based on the query constraints of the penetration query request, perform a hierarchical penetration path search on the multi-source heterogeneous association graph, and determine the set of penetration control strategies for the enterprise information platform based on the node level and business type of each searched work path. Step 3: Based on the user-initiated penetrating work request, access each level of nodes on the work path in sequence. When accessing each node, call the corresponding control policy in the penetrating control policy set for compliance verification. Summarize the execution logs of all nodes to obtain the full-level control result report of the enterprise information platform.

2. The method for generating comprehensive management and control based on multi-source heterogeneous data as described in claim 1, characterized in that, The method involves constructing a multi-source heterogeneous association graph for the enterprise information platform, using data entities and entity attribute information from multiple heterogeneous data sources as nodes and semantic similarity relationships between data entities as edges. This includes: The raw data extracted from the heterogeneous data source is parsed and identified to obtain the data entities and corresponding entity attribute information in the raw data, and a temporary entity identifier is generated for each data entity. Based on temporary entity identifiers, calculate the semantic similarity between data entities and their corresponding entity attribute information; When the semantic similarity exceeds a preset threshold, it is determined that there is a relationship between the two corresponding data entities; Using data entities as nodes and relationships as edges, a multi-source heterogeneous relationship graph is constructed for the enterprise information platform.

3. The method for generating comprehensive management and control based on multi-source heterogeneous data as described in claim 2, characterized in that... The step of parsing and identifying the raw data extracted from heterogeneous data sources includes: Obtain the full-level organizational structure list of the enterprise information platform. The full-level organizational structure list includes the names of all nodes from the highest management level to the grassroots operating units and the reporting relationships between nodes. The business department identifier and business level identifier of the original data recorded in each heterogeneous data source will be read and bound to the corresponding organizational node to form an organizational node mapping table for the heterogeneous data source. Based on the organization node mapping table, query the data source type and data source identifier of the entity in each heterogeneous data source to obtain the node path and associated entities of the organization node. The associated entities include regulatory object entities, business matter entities, regulatory indicator entities and regulatory rule entities. When the same organization node is associated with entities from multiple heterogeneous data sources, all entities belonging to the organization node are merged into a comprehensive entity set of the organization node; The original source tags of each source entity are retained within the comprehensive entity set to distinguish the data differences in the description of the same organizational node by different business platforms.

4. The method for generating comprehensive management and control based on multi-source heterogeneous data as described in claim 2, characterized in that, The formula for calculating semantic similarity is as follows: ； In the formula, To indicate the first The data entity and the first Semantic similarity between data entities For data entities, This represents the total number of key attribute types in the entity attribute information. For the first The preset weight coefficients of various attributes in semantic similarity calculation It is a cosine function. For the first The first data entity The first attribute vector obtained after vectorizing the attribute values ​​is the first attribute value. One portion, For the first The first data entity The second attribute vector obtained after the attribute value is vectorized One portion, For the first The set of contextual identifiers for each data entity appearing in the enterprise information platform. For the first The set of contextual identifiers for each data entity appearing in the enterprise information platform.

5. The method for generating comprehensive management and control based on multi-source heterogeneous data as described in claim 1, characterized in that, The step of performing a hierarchical penetration path search on the multi-source heterogeneous association graph based on the query constraints of the penetration query request includes: Accepting penetrating query requests initiated by users through the enterprise-commission linkage system, the parsed target organization node identifier, penetration depth level, and the type of business matter to be queried are aggregated into query constraints for a multi-source heterogeneous association graph; Based on query constraints, starting from the target organization node identifier, the corresponding starting node is located in the multi-source heterogeneous association graph, and the node attribute information of the starting node in the graph is obtained. Based on the penetration depth level, starting from the starting node, perform a breadth-first traversal along the reporting membership edges in the graph to the next lower level node.

6. The method for generating comprehensive management and control based on multi-source heterogeneous data as described in claim 5, characterized in that... The breadth-first traversal includes: When traversing each level, the nodes visited at the current level are recorded as path nodes, until the deepest level specified by the penetration depth level number is reached or the current level has no lower-level nodes, then the traversal stops. During the traversal, based on the type of business matter to be queried, nodes in the entity set associated with each path node that contain entity information matching the type of business matter are retained as retained nodes. Combine the reporting hierarchy edges between nodes into one or more complete hierarchical penetration paths.

7. The method for generating comprehensive management and control based on multi-source heterogeneous data as described in claim 6, characterized in that, The process involves determining a set of penetrating control strategies for the enterprise information platform based on the node level and business type of each work path found through the search, including: Based on the organizational hierarchy attribute of the hierarchical penetration path, each node is matched with the permission control template corresponding to the current level. The permission control template defines the data viewing scope, data modification permissions, and data reporting requirements allowed for the corresponding level node. Based on the type of business matter, a corresponding rule control template is matched for each node. The rule control template includes threshold comparison rules, data integrity verification rules, and reporting timeliness verification rules for business matters under the commission-enterprise linkage system. The permission control template and the rule control template will be used together as the single-node control strategy in this penetration query. Use the hierarchical penetration path as the working path; The control strategies for each node are statistically summarized according to the order of nodes in the work path, resulting in a set of penetrating control strategies for the enterprise information platform.

8. The method for generating comprehensive management and control based on multi-source heterogeneous data as described in claim 1, characterized in that, The process involves sequentially accessing nodes at each level of the work path based on user-initiated penetration-based work requests. When accessing each node, the corresponding control policy from the penetration-based control policy set is invoked for compliance verification, including: Extract the target work path identifier and the type of operation to be performed from the user-initiated through-work request, and determine that the target work path identifier points to an already generated work path; According to the order of the corresponding nodes in the work path, the process starts from the first node and proceeds downwards. When a node is reached, the subsequent access is paused and the process switches to the strategy invocation phase. Based on the node identifier of the corresponding node and the type of operation to be executed, retrieve the matching control policy from the set of penetrating control policies; Execute the retrieved control policy, complete the compliance verification of the corresponding node, and decide whether to continue accessing the next node on the path or terminate the entire work request based on the verification results and the feedback data after execution.

9. The method for generating comprehensive management and control based on multi-source heterogeneous data as described in claim 8, characterized in that... The completion of compliance verification for the current node includes: Read the data access permission rules stored in the metadata of the corresponding node. The data access permission rules include a list of user roles that are allowed to access the corresponding node and a list of operations that are allowed to be performed. Based on the user's role in the penetrating work request, the user role is matched with the list of user roles. If the match fails, the verification is directly determined to be unsuccessful and subsequent access is denied. After a successful role match, further check whether the type of operation to be performed is included in the list of allowed operations. If it is not included, the verification fails and the operation of the current node is terminated. If both the role and the operation type match successfully, a compliance record containing the current node identifier, user identifier, operation timestamp, and operation type will be generated according to the operation record requirements of the penetration-type work request. The compliance record and the post-execution feedback data will be temporarily stored in the verification result cache. After all corresponding nodes have been accessed, all compliance records in the cache and post-execution feedback data are aggregated into the execution log of the penetrating work request.

10. The method for generating comprehensive management and control based on multi-source heterogeneous data as described in claim 9, characterized in that... The aggregation of execution logs from all nodes yields a comprehensive control result report for the enterprise information platform, including: According to the hierarchical order of the nodes in the working path, sort and merge the execution logs corresponding to all nodes to obtain the log sequence of the nodes; Based on the verification result status and post-execution feedback data of each execution log in the log sequence, count the number of nodes that passed verification and the number of nodes that failed verification on the working path, and mark the specific level position of the nodes that failed verification. The log sequence, statistical results, and marked non-passing node information are assembled into a template to generate a full-level control result report for the enterprise information platform.

Images