CN-121979840-A - Knowledge graph-based archive multidimensional element management method and system

CN121979840ACN 121979840 ACN121979840 ACN 121979840ACN-121979840-A

Abstract

The invention discloses a method and a system for managing multidimensional file elements based on a knowledge graph, which relate to the technical field of file information management and comprise the steps of carrying out element identification and relation extraction on file information resources to obtain an initial association relation between a file entity set and an entity, and constructing a file knowledge graph based on the initial association relation; the method and the system can comprehensively extract the event characteristics comprising direct operation information, local context and global association influence, and then generate the association transition contract covering the change basis, the operation instruction and the verification condition by combining the current state of the atlas, thereby realizing self-adaptive response to entity association dynamic evolution and ensuring that the atlas can reflect real-time and accurate element relation change in the real world.

Inventors

CHENG JIE

Assignees

烟台市博物馆(烟台市考古研究院、胶东革命纪念馆、烟台山文物管理服务中心)

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (10)

1. The archive multidimensional element management method based on the knowledge graph is characterized by comprising the following steps of: S1, carrying out element identification and relation extraction on file information resources to obtain a file entity set and an initial association relation between the entities, and constructing a file knowledge graph based on the initial association relation; S2, monitoring file change events in a file knowledge graph, and extracting event characteristics of the change events; s3, carrying out fusion analysis based on the event characteristics and the current state of the archive knowledge graph to generate an associated transition contract comprising a change basis, an operation instruction and a verification condition; s4, verifying and executing an operation instruction in the associated transition contract, updating the archive knowledge graph, and generating a local graph snapshot corresponding to the associated transition contract; s5, generating a tracing analysis result of the association transition when the query request is executed based on the association transition contract and the local map snapshot.
2. The method for multidimensional file element management based on knowledge graph as claimed in claim 1, wherein said performing element recognition and relationship extraction on file information resources to obtain an initial association relationship between a set of file entities and the entities comprises: analyzing the structured data table in the archive information resource, extracting a first part of entities forming an archive entity set, and generating a first association relationship among the first part of entities according to the constraint relationship among the data tables; Analyzing unstructured documents in the archive information resource, extracting second partial entities forming an archive entity set, and generating a second association relationship between the second partial entities according to text semantic analysis; And combining the first association relationship and the second association relationship to obtain an initial association relationship between the entities.
3. The method for multidimensional file element management based on a knowledge graph of claim 1, wherein monitoring file change events in the knowledge graph of the file comprises: Monitoring and analyzing the data flow acting on the archive knowledge graph, and identifying the operation behavior of the entity or the relation to obtain an original change event; coupling analysis is carried out on the original change event and the complete state of the archive knowledge graph when the event occurs, so as to obtain event basic characteristics comprising direct operation objects, operation types and local graph contexts; In a global association network of the archive knowledge graph, carrying out influence deduction on the change indicated by the event basic feature, and identifying an affected association path and a remote entity to obtain an association influence feature of the event; And integrating the event basic features and the associated influence features to form event features.
4. The method for multidimensional file element management based on knowledge graph as claimed in claim 1, wherein the generating the associated transition contract including the change basis, the operation instruction and the verification condition based on the fusion analysis of the event feature and the current state of the file knowledge graph comprises: Comparing and analyzing event characteristics with the historical states and the association strength of the affected entities in the archive knowledge graph to obtain compliance judgment and influence evaluation of the change event, wherein the compliance judgment and the influence evaluation are used as change basis; mapping entity attribute updating and relation adjusting steps required to be executed for realizing the change in the archive knowledge graph according to the change basis, and taking the entity attribute updating and relation adjusting steps as operation instructions; Deriving constraint rules for ensuring semantic consistency and structural integrity of the changed patterns based on operation types and influence ranges in the event characteristics and by combining the associated network structure of the archive knowledge patterns, and taking the constraint rules as verification conditions; And packaging the change basis, the operation instruction and the verification condition according to a preset contract structure to obtain the associated transition contract.
5. The method for multidimensional file element management based on knowledge graph as claimed in claim 4, wherein the deriving constraint rules for ensuring semantic consistency and structural integrity of the changed graph by combining the operation type and influence range in the event feature and the associated network structure of the knowledge graph as verification conditions comprises: In the archive knowledge graph, performing influence simulation on the change operation indicated by the event characteristics, and deducing an entity state chain change path possibly triggered by the change operation in the associated network structure; Consistency comparison is carried out on the entity history version related in the chain change path and the current association relation, and association nodes and association edges with conflict risks are identified; And constructing constraint rules for conflict detection and consistency verification during contract execution aiming at the associated nodes and associated edges with conflict risks, and forming verification conditions.
6. The method for multidimensional file element management based on knowledge graph as claimed in claim 1, wherein verifying and executing the operation instruction in the associated transition contract updates the knowledge graph of the file, and simultaneously generates a local graph snapshot corresponding to the associated transition contract, comprising: Based on the verification conditions in the associated transition contracts, carrying out compliance verification on the current state of the archive knowledge graph, and executing change operation on the archive knowledge graph based on the operation instruction; extracting state information of affected entities and relations before and after the change in the process of executing the change operation; and packaging the extracted state information and the unique identifier of the associated transition contract to form a local map snapshot.
7. The knowledge-based archive multidimensional element management method as recited in claim 6, wherein the encapsulating the extracted state information with the unique identifier of the associated transition contract to form a local atlas snapshot includes: Positioning a directly modified entity set and a directly modified relation set in the archive knowledge graph based on the operation instruction, and taking the entity set and the relation set as a core change set; In a global association network of the archive knowledge graph, carrying out one-time association expansion on the core change set, and identifying all directly associated but unmodified entities and relationships as an association context set; and combining the state before and after the change of each element in the core change set, the current state of the associated context set and the unique identifier of the associated transition contract, and packaging the combination into a local map snapshot.
8. The knowledge-graph-based archive multidimensional element management method as recited in claim 1, wherein the generating a trace-source analysis result of the association transition when the query request is executed based on the association transition contract and the local graph snapshot comprises: Analyzing the received query request, and locating a target entity or a target relation to be traced from the received query request; Based on a target entity or a target relation, matching and searching are carried out in a set of associated transition contracts and a set of local map snapshots, and all related contracts and snapshots are screened out to form a traceable evidence chain material; Comparing and analyzing the local map states before and after the execution of the operation instruction in the tracing evidence chain material to obtain a direct state transition path of the target entity or the target relation; And carrying out association integration on the direct state transition path and the corresponding change basis and verification condition to form a tracing analysis result of association transition.
9. The knowledge-graph-based archive multidimensional element management method as claimed in claim 8, wherein the comparing and analyzing the local graph states before and after the execution of the operation instruction in the trace-source evidence chain material to obtain the direct state transition path of the target entity or the target relationship comprises: For a target entity or a target relation, extracting a complete association context state before modification and a complete association context state after modification from a local map snapshot where the target entity or the target relation is located; Performing difference calculation on the complete associated context state before and after the change, and identifying all associated edges and associated nodes which change due to the execution of the contract; and sequencing and connecting all the identified changes according to the logic sequence and the dependency relationship of the changes in the global network of the archive knowledge graph to obtain a direct state transition path.
10. A knowledge-graph-based archive multidimensional element management system for implementing the knowledge-graph-based archive multidimensional element management method as recited in any one of claims 1 to 9, the system comprising: The map construction module is used for carrying out element identification and relation extraction on the archive information resource to obtain an archive entity set and an initial association relation between the entities, and constructing an archive knowledge map based on the archive entity set and the initial association relation; The feature extraction module is used for monitoring file change events in the file knowledge graph and extracting event features of the change events; The map analysis module is used for carrying out fusion analysis based on the event characteristics and the current state of the archive knowledge map to generate an associated transition contract comprising a change basis, an operation instruction and a verification condition; the map snapshot generating module is used for verifying and executing the operation instruction in the associated transition contract, updating the archive knowledge map and generating a local map snapshot corresponding to the associated transition contract; and the traceability analysis module is used for generating traceability analysis results of the association transition when the query request is executed based on the association transition contract and the local map snapshot.

Description

Knowledge graph-based archive multidimensional element management method and system Technical Field The invention relates to the technical field of archive information management, in particular to a knowledge-graph-based archive multidimensional element management method and system. Background At present, a method for managing file multidimensional elements by using a knowledge graph generally comprises the steps of identifying and extracting file entities, then constructing association among the entities according to preset rules or simple co-occurrence relations, and finally using the formed knowledge graph for retrieval, visualization or simple analysis. In the prior art, in some major engineering project archive management, defects exist, because the participants, responsibility attributions and key places of projects may be changed along with the progress of project stages, once the construction of the existing static knowledge graph is completed, the internal association relationship is solidified, and the system cannot automatically sense the change of the real world element relationship reflected behind the archive record, so that information lag and even errors occur in applications such as archive association retrieval, responsibility tracing and experience reference based on the old graph. Disclosure of Invention The invention provides a knowledge graph-based archive multidimensional element management method and a knowledge graph-based archive multidimensional element management system, which are used for solving the problems in the background technology. In order to achieve the above object, the method for managing archive multidimensional elements based on knowledge graph provided by the present invention comprises: S1, carrying out element identification and relation extraction on file information resources to obtain a file entity set and an initial association relation between the entities, and constructing a file knowledge graph based on the initial association relation; S2, monitoring file change events in a file knowledge graph, and extracting event characteristics of the change events; s3, carrying out fusion analysis based on the event characteristics and the current state of the archive knowledge graph to generate an associated transition contract comprising a change basis, an operation instruction and a verification condition; s4, verifying and executing an operation instruction in the associated transition contract, updating the archive knowledge graph, and generating a local graph snapshot corresponding to the associated transition contract; s5, generating a tracing analysis result of the association transition when the query request is executed based on the association transition contract and the local map snapshot. Preferably, the element identification and relationship extraction of the archive information resource to obtain an archive entity set and an initial association relationship between the entities includes: analyzing the structured data table in the archive information resource, extracting a first part of entities forming an archive entity set, and generating a first association relationship among the first part of entities according to the constraint relationship among the data tables; Analyzing unstructured documents in the archive information resource, extracting second partial entities forming an archive entity set, and generating a second association relationship between the second partial entities according to text semantic analysis; And combining the first association relationship and the second association relationship to obtain an initial association relationship between the entities. Preferably, the monitoring the archive change event in the archive knowledge graph includes: Monitoring and analyzing the data flow acting on the archive knowledge graph, and identifying the operation behavior of the entity or the relation to obtain an original change event; coupling analysis is carried out on the original change event and the complete state of the archive knowledge graph when the event occurs, so as to obtain event basic characteristics comprising direct operation objects, operation types and local graph contexts; In a global association network of the archive knowledge graph, carrying out influence deduction on the change indicated by the event basic feature, and identifying an affected association path and a remote entity to obtain an association influence feature of the event; And integrating the event basic features and the associated influence features to form event features. Preferably, the generating the association transition contract including the change basis, the operation instruction and the verification condition based on the fusion analysis of the event feature and the current state of the archive knowledge graph includes: Comparing and analyzing event characteristics with the historical states and the association strength of the affected entities in