Search

CN-121387855-B - Knowledge graph-based ERP data migration and cleaning automation method and system

CN121387855BCN 121387855 BCN121387855 BCN 121387855BCN-121387855-B

Abstract

The invention provides an ERP data migration and cleaning automation method and system based on a knowledge graph, and relates to the technical field of enterprise data management, wherein the method comprises the steps of constructing the knowledge graph of the ERP field according to business requirements and rules; the method comprises the steps of carrying out semantic field matching and conflict detection on a source system and a target system by a knowledge graph reasoning engine, executing automatic data cleaning, migrating data according to entity association relation and carrying out service consistency verification, generating a complete report containing migration success rate, cleaning statistics, abnormal lists and processing suggestions, and generating six modules including knowledge graph construction, system docking and analysis, semantic matching and conflict detection, automatic data cleaning, data migration and verification and report generation.

Inventors

  • LIU GUANGJUN

Assignees

  • 北京浩太同益科技发展有限公司

Dates

Publication Date
20260512
Application Date
20251029

Claims (10)

  1. 1. The ERP data migration and cleaning automation method based on the knowledge graph is characterized by comprising the following steps of: step 1, constructing an ERP domain knowledge graph comprising an ERP business entity, an entity association relationship and a business constraint rule according to business domain requirements of an ERP system and preset business entity rules; Step 2, respectively interfacing a source ERP system and a target ERP system according to the knowledge graph of the ERP field, analyzing the data structure of data to be migrated in the source system and outputting source data structure information, analyzing the data receiving specification of the target system and outputting target data structure information; Step 3, according to the source data structure information, the target data structure information and the entity association relation and business constraint rules in the knowledge graph of the ERP field, carrying out semantic field matching through a knowledge graph reasoning engine, outputting a field mapping relation, and detecting the difference between the source data structure and the target data structure to output conflict information; Step 4, executing automatic data cleaning operation according to the field mapping relation, conflict information and business constraint rules, and outputting cleaned data; Step 5, migrating the data to a target ERP system according to the cleaned data, the field mapping relation and the entity association relation, obtaining a data migration result, and performing service consistency check on the migrated data based on a knowledge graph traversal service entity association chain to obtain a check result; And 6, generating an ERP data migration cleaning report comprising migration success rate, cleaning processing statistical information, an abnormal record list and processing suggestions according to the data migration result and the verification result.
  2. 2. The knowledge graph-based ERP data migration and cleaning automation method of claim 1, wherein the step 1 of constructing the ERP domain knowledge graph comprises extracting entities and relations from standard data table structures, industry specifications and enterprise internal data ledgers based on a BERT entity identification model, and storing the entities and relations by adopting a Neo4j graph database after manual verification, and iteratively updating the relationships in each quarter.
  3. 3. The knowledge-based ERP data migration and cleansing automation method of claim 1, wherein the interfacing the source and target systems in step 2 comprises determining an interfacing interface specification based on the knowledge-based profile, configuring interface parameters and establishing a bi-directional communication link through connectivity verification rule testing.
  4. 4. The knowledge-graph-based ERP data migration and cleaning automation method according to claim 1, wherein in the step 3, semantic level field matching comprises calculating field name semantic similarity through a Word2Vec model and weighting and calculating matching confidence by combining data types, lengths and mandatory properties, conflict information comprises field structure conflict and business constraint conflict, the field structure conflict comprises inconsistent data types, lengths and major foreign key properties, and the business constraint conflict comprises data formats, numerical verification and association integrity violations.
  5. 5. The knowledge-graph-based ERP data migration and cleansing automation method of claim 1, wherein the automated data cleansing operation in step 4 performs corresponding processing for different conflict types, including data type conversion, field length expansion, data format completion, numerical correction, and association information completion.
  6. 6. The knowledge-graph-based ERP data migration and cleaning automation method according to claim 1, wherein in step 5, data migration is performed according to an entity migration priority table, priorities are divided according to entity association relations, a referenced entity is preferentially migrated over a referenced entity, and business consistency verification is performed by traversing an entity association chain in a knowledge graph to verify association field matching and business constraint compliance.
  7. 7. A system for the knowledge-based ERP data migration and cleansing automation method of any one of claims 1-6, the system comprising: the knowledge graph construction module is used for constructing an ERP field knowledge graph comprising an ERP business entity, an entity association relationship and a business constraint rule; The system docking and analyzing module is used for docking the source and target ERP systems, analyzing the data structure and outputting the source and target data structure information; The semantic matching and conflict detection module is used for carrying out semantic level field matching and outputting a field mapping relation through the knowledge graph reasoning engine, detecting structural differences and outputting conflict information; the automatic data cleaning module is used for executing automatic data cleaning operation and outputting cleaned data; The data migration and verification module is used for migrating data to a target system and obtaining migration results, and carrying out service consistency verification and obtaining verification results; And the report generation module is used for generating a data migration cleaning report containing migration success rate, cleaning statistics, an abnormal list and processing suggestions.
  8. 8. The system of claim 7, wherein the knowledge graph construction module extracts entities and relationships using a BERT-based entity recognition model and stores and updates knowledge graphs via a Neo4j graph database, and the semantic matching and collision detection module integrates a Word2Vec model and weighted computation logic for field semantic similarity and attribute matching computation.
  9. 9. The system of claim 7, wherein the data migration and verification module supports performing migration in entity migration priority table order and implementing service consistency verification through association chain traversal, and wherein the automated data cleansing module comprises a rule execution unit for parsing and executing service constraint rules defined based on SWRL language to handle different types of conflicts.
  10. 10. An electronic device comprising a processor and a memory, wherein the memory stores a computer program which, when executed by the processor, implements the method of any of claims 1-6.

Description

Knowledge graph-based ERP data migration and cleaning automation method and system Technical Field The invention relates to the technical field of enterprise data management, in particular to an ERP data migration and cleaning automation method and system based on a knowledge graph. Background In the process of enterprise digital operation, an Enterprise Resource Planning (ERP) system is used as a key platform for integrating core business data such as enterprise production, purchasing, inventory, finance, human resources and the like, and the stability and the data quality of the ERP system directly determine the continuity of enterprise business processes and the accuracy of decisions. With the expansion of enterprise business scale, the transformation of a created environment, the version upgrading of an ERP system or the integration of multiple systems, the transfer and cleaning of ERP data become the core links in the digitized transformation of the enterprise, the requirements for transferring business data in a historical system or a heterogeneous system to a target ERP system are more and more increased, but the prior art is focused on the local functions of the ERP system such as report management, or only solves the single links of general data transfer such as database adaptation, or the special data scenes of non-ERP such as streaming data analysis, such as patent 202410913830.7 discloses an online custom report management method and system combined with the ERP system, has obvious scene limitation, does not cover the ERP data transfer scene, lacks data cleaning capability, does not introduce knowledge modeling means in the fields such as knowledge graph, and cannot recognize the business correlation of field dependence relationship such as sales order report and client main data report after ERP report data, patent 202411771321.1 proposes an automatic matching method and system for data transfer, such as patent 202410913830.7 discloses an automatic matching method, a device for data transfer, computer equipment and a storage device for non-ERP data analysis, has no correlation with the prior art, and has no correlation with the knowledge modeling means of the prior art, and has no correlation with the service data has a great knowledge-based on the service correlation, and has no correlation with the prior art, and has no correlation with the service data cleaning capability, the accuracy and consistency of the data cannot be guaranteed from the semantic level. In view of the foregoing, there is a need for an ERP data migration and cleaning automation method and system based on knowledge maps to solve at least the above-mentioned shortcomings. Disclosure of Invention The invention aims to provide an ERP data migration and cleaning automation method and system based on a knowledge graph, which are used for solving the problems of low efficiency, high cost, poor data quality, high business interruption risk and the like faced by enterprises in the ERP data migration and cleaning process in the prior art, and the specific technical scheme is as follows: The invention provides an ERP data migration and cleaning automation method based on a knowledge graph, which comprises the following steps: step 1, constructing an ERP domain knowledge graph comprising an ERP business entity, an entity association relationship and a business constraint rule according to business domain requirements of an ERP system and preset business entity rules; Step 2, respectively interfacing a source ERP system and a target ERP system according to the knowledge graph of the ERP field, analyzing the data structure of data to be migrated in the source system and outputting source data structure information, analyzing the data receiving specification of the target system and outputting target data structure information; Step 3, according to the source data structure information, the target data structure information and the entity association relation and business constraint rules in the knowledge graph of the ERP field, carrying out semantic field matching through a knowledge graph reasoning engine, outputting a field mapping relation, and detecting the difference between the source data structure and the target data structure to output conflict information; Step 4, executing automatic data cleaning operation according to the field mapping relation, conflict information and business constraint rules, and outputting cleaned data; Step 5, migrating the data to a target ERP system according to the cleaned data, the field mapping relation and the entity association relation, obtaining a data migration result, and performing service consistency check on the migrated data based on a knowledge graph traversal service entity association chain to obtain a check result; And 6, generating an ERP data migration cleaning report comprising migration success rate, cleaning processing statistical information, an abnormal record list and proce