An Automatic Reconciliation System for Enterprise Data and Its Implementation Method

Abstract

The present invention relates to the technical field of data management, and particularly to an automatic reconciliation system for enterprise data and an implementation method thereof. The method includes the following steps: obtaining multi-source enterprise bill data; performing data preprocessing on the multi-source enterprise bill data to generate standard multi-source enterprise bill data; decomposing the enterprise bill addresses of the standard multi-source enterprise bill data to generate bill type classification data; archiving and storing the bill type classification data in a database to generate a normalized enterprise storage data set; performing storage table association on the normalized enterprise storage data set to generate a time-series associated storage table; and performing step-by-step initial bill reconciliation on the time-series associated storage table to generate step-by-step initial bill reconciliation data. The present invention improves the automation level and accuracy of the reconciliation process through systematic data preprocessing, normalized storage, automatic initial verification and correction, intelligent anomaly detection, and rule generation.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to an enterprise data automatic reconciliation system and an implementation method thereof. Background Art

[0002] Initially, enterprise reconciliation mainly relied on manual operations, involving a large number of paper records and manual verification. This method is not only time-consuming and laborious but also prone to human errors. With the development of information technology, enterprises gradually introduced computer systems to store and manage financial data. At this stage, although the degree of automation has increased, data reconciliation still mainly relies on regular data aggregation and manual comparison, and the efficiency and accuracy are limited. Entering the 21st century, significant progress has been made in data reconciliation technology. Enterprises began to adopt database management systems and ERP (Enterprise Resource Planning) systems, which can centrally process and integrate financial data from different departments. Nevertheless, the automatic reconciliation function is still limited, mainly focusing on simple data comparison and basic anomaly detection. In recent years, with the development of big data technology and artificial intelligence, the technical level of enterprise data automatic reconciliation has been significantly improved. Machine learning algorithms and intelligent data analysis tools have been introduced, which can automatically identify complex data anomalies and inconsistencies. These technologies not only improve the accuracy of reconciliation but also significantly reduce the time and labor costs required. However, currently, traditional bill reconciliation does not fully consider the classification and standardized storage of bills, resulting in complex subsequent processing and low efficiency. At the same time, the marking and warning mechanism for abnormal bills is not perfect, making it difficult to discover and process abnormal bills in a timely manner, thereby leading to a low level of automation and accuracy in the reconciliation process. Summary of the Invention

[0003] Based on this, it is necessary to provide an enterprise data automatic reconciliation system and an implementation method thereof to solve at least one of the above technical problems.

[0004] To achieve the above object, an implementation method for enterprise data automatic reconciliation, the method includes the following steps:

[0005] Step S1: Obtain multi-source enterprise bill data; perform data preprocessing on the multi-source enterprise bill data to generate standard multi-source enterprise bill data; perform enterprise bill address decomposition on the standard multi-source enterprise bill data to generate bill type classification data; perform classified database archiving storage on the bill type classification data to generate a standardized enterprise storage data set;

[0006] Step S2: Perform storage table association on the standardized enterprise storage data set to generate a time-series association storage table; perform step-by-step initial bill reconciliation on the time-series association storage table to generate step-by-step initial bill reconciliation data; perform differential ledger correction on the step-by-step initial bill reconciliation data to generate a time-series association storage correction table;

[0007] Step S3: Conduct a difference review on the time-series correlation storage correction table to generate review difference result data; based on the review difference result data, mark abnormal bills on the time-series correlation storage correction table to generate abnormal bill mark data; issue an early warning for the reasons of abnormalities for the abnormal bill mark data to generate abnormal reconciliation record early warning data; verify the abnormal bill mark data through the abnormal reconciliation record early warning data to generate abnormal reconciliation record early warning feedback data.

[0008] Step S4: Generate an enterprise automatic reconciliation rule engine by generating reconciliation rules for the abnormal reconciliation record early warning feedback data and the step-by-step bill preliminary reconciliation data; based on the enterprise automatic reconciliation rule engine, conduct automatic reconciliation and storage of multi-source enterprise bill data to generate enterprise bill block storage data, so as to execute the enterprise data automatic reconciliation storage operation.

[0009] Through preprocessing and standardization of multi-source enterprise bill data, the present invention ensures that all data formats are unified and generates a standardized enterprise storage data set, which not only improves the accuracy of the data, but also makes the subsequent bill processing more efficient and consistent. Through step-by-step bill preliminary reconciliation and difference ledger correction, a time-series correlation storage correction table is generated, automating the preliminary reconciliation and difference correction processes of the bills, greatly reducing manual operations and human errors, and improving the efficiency and accuracy of reconciliation. Conducting a difference review on the time-series correlation storage correction table to generate review difference result data, and based on these data, marking and warning of abnormal bills can timely detect abnormal bills and generate abnormal reconciliation record early warning data, so as to quickly respond to and handle potential financial problems. By generating an enterprise automatic reconciliation rule engine from the abnormal reconciliation record early warning feedback data and the step-by-step bill preliminary reconciliation data, the reconciliation rules can be dynamically generated and adjusted to ensure that the reconciliation process is more intelligent and personalized, improving the accuracy and adaptability of automatic reconciliation. Storing the reconciled data as enterprise bill block storage data realizes the structured storage and management of the data, improves the retrieval efficiency and storage space utilization rate of the data, and supports the efficient processing and long-term preservation of large-scale enterprise bill data. Therefore, through systematic data preprocessing, standardized storage, automated preliminary verification and correction, intelligent anomaly detection and rule generation, the present invention improves the automation level and accuracy of the reconciliation process.

[0010] Preferably, step S1 includes the following steps:

[0011] Step S11: Obtain multi-source enterprise bill data;

[0012] Step S12: Conduct data preprocessing on the multi-source enterprise bill data to generate standard multi-source enterprise bill data, where the data preprocessing includes data cleaning, filling of missing data values, processing of data outliers, and data standardization.

[0013] Step S13: Track the enterprise bill address for the standard multi-source enterprise bill data to generate enterprise bill address tracking data; decompose the standard multi-source enterprise bill data according to the enterprise bill address tracking data to generate bill type classification data, where the bill type classification data includes bank statement data, enterprise financial record data, and customer transaction record data;

[0014] Step S14: Archive and store based on the bank statement data, enterprise financial record data, and customer transaction record data into separate databases to generate a normalized enterprise storage data set.

[0015] The present invention ensures data consistency and integrity, improves data quality and usability by performing data cleaning, missing value filling, outlier handling, and standardization on multi-source enterprise bill data. Tracking the enterprise bill address helps to accurately locate and manage the bill source. Decomposing the standard multi-source enterprise bill data into different bill types (bank statements, enterprise financial records, and customer transaction records) helps with classified management and processing, improving the efficiency and accuracy of data analysis. Archiving and storing the data according to the bill type to generate a normalized enterprise storage data set helps with the structured management and long-term preservation of data, improving the retrieval and usage efficiency of data. The normalized and classified storage data set is convenient for subsequent financial analysis, risk assessment, and decision support, enhancing the utilization value of data and the accuracy of business decisions.

[0016] Preferably, step S13 includes the following steps:

[0017] Step S131: Parse the multi-source data format of the standard multi-source enterprise bill data to generate multi-source format parsing data;

[0018] Step S132: Extract the bill IP information from the standard multi-source enterprise bill data according to the multi-source format parsing data to obtain multi-source bill IP information data;

[0019] Step S133: Verify the external address for the multi-source bill IP information data to generate multi-source bill address verification data; encode the longitude and latitude coordinates for the multi-source bill IP information data through the multi-source bill address verification data to generate multi-source bill longitude and latitude information data; perform address tracking and matching on the multi-source bill longitude and latitude information data through GIS technology to generate enterprise bill address tracking data;

[0020] Step S134: Decompose the standard multi-source enterprise bill data according to the enterprise bill address tracking data to generate bill type classification data, where the bill type classification data includes bank statement data, enterprise financial record data, and customer transaction record data.

[0021] The present invention ensures the compatibility and consistency of data among different sources by parsing the standard multi-source enterprise bill data in multiple data formats, generating multi-source format parsed data, which helps the smooth progress of subsequent processing. Extract the bill IP information from the standard multi-source enterprise bill data, and generate multi-source bill address verification data through external address verification, ensuring the accuracy and reliability of the bill IP information. By encoding the longitude and latitude coordinates of the multi-source bill IP information data, generating multi-source bill longitude and latitude information data, and then combining with GIS technology for address tracking and matching, it helps to accurately locate the address information of enterprise bills, improving the accuracy and efficiency of address tracking. Decompose the standard multi-source enterprise bill data according to the enterprise bill address tracking data, generating bill type classification data, including bank statement data, enterprise financial record data, and customer transaction record data, which helps the classified management and analysis of bill data, improving the accuracy and efficiency of data processing. Through steps such as multi-source data format parsing, IP information extraction and verification, longitude and latitude encoding and address matching, the accuracy and reliability of enterprise bill data in different processing stages are ensured, reducing the possibility of data errors and mistakes. The classified bill data is more convenient for financial analysis, risk assessment, and business decision support, helping enterprises better understand and utilize bill data, and enhancing the commercial value of data.

[0022] Preferably, step S134 includes the following steps:

[0023] Step S1341: Screen the common addresses from the enterprise bill address tracking data to obtain common address bill data and non-common address bill data;

[0024] Step S1342: Conduct a first type tendency analysis on the standard multi-source enterprise bill data through the common address bill data to generate first type tendency data, where the first type tendency analysis includes bank reconciliation analysis and enterprise financial analysis; conduct a second type tendency analysis on the standard multi-source enterprise bill data through the non-common address bill data to generate second type tendency data, where the second type tendency analysis includes customer transaction record analysis;

[0025] Step S1343: Extract the bill content features from the standard multi-source enterprise bill data to obtain bill content feature data; decompose the bill content feature data according to the first type tendency data and the second type tendency data to generate bill type classification data, where the bill type classification data includes bank statement data, enterprise financial record data, and customer transaction record data.

[0026] The present invention screens for common addresses from the enterprise bill address tracking data, differentiating between bill data for common addresses and bill data for non-common addresses, which helps to conduct more refined analysis and management for different address types, improving the accuracy and efficiency of data processing. Bank reconciliation analysis and enterprise financial analysis are performed on the standard multi-source enterprise bill data using the bill data for common addresses to generate the first type of tendency data. This helps to understand the financial activities and bank transactions at the main operating addresses of the enterprise, supporting the financial management and decision-making of the enterprise. Customer transaction record analysis is performed on the standard multi-source enterprise bill data using the bill data for non-common addresses to generate the second type of tendency data. This helps to identify the customer transaction activities of the enterprise at secondary addresses, providing a more comprehensive analysis of business activities. Feature extraction of the bill content is performed on the standard multi-source enterprise bill data to generate bill content feature data. By extracting the features of the bill content, the specific content of the bill data can be analyzed more deeply, enhancing the meticulousness and accuracy of data analysis. The bill content feature data is decomposed into bank statement data, enterprise financial record data, and customer transaction record data according to the first type of tendency data and the second type of tendency data. This process ensures the accuracy of bill data classification, facilitating subsequent analysis and processing. By differentiating between common addresses and non-common addresses and their tendency analysis, bill data can be processed and managed more targeted, improving the precision and efficiency of data processing. The generation of bill type classification data helps the enterprise to conduct more meticulous financial analysis, risk assessment, and decision support, enhancing the utilization value of the enterprise's bill data and the accuracy of business decisions.

[0027] Preferably, step S2 includes the following steps:

[0028] Step S21: Perform time-series amount transaction analysis on the normalized enterprise storage dataset to generate time-series amount transaction data;

[0029] Step S22: Perform storage table association on the normalized enterprise storage dataset using the time-series amount transaction data to generate a time-series associated storage table, where the time-series associated storage table includes a bank-enterprise amount association table and an enterprise-customer amount association table;

[0030] Step S23: Perform step-by-step initial bill reconciliation on the bank-enterprise amount association table and the enterprise-customer amount association table to generate step-by-step initial bill reconciliation data; perform difference identification on the step-by-step initial bill reconciliation data to generate step-by-step bill difference identification data;

[0031] Step S24: Perform ledger correction on the bank-enterprise amount association table and the enterprise-customer amount association table according to the step-by-step bill difference identification data to generate a time-series associated storage correction table.

[0032] The present invention generates time - series amount transaction data through time - series amount transaction analysis of the standardized enterprise storage data set. This helps to identify the transaction patterns and trends of the enterprise in different time periods, providing a more comprehensive view of financial activities. Through the time - series amount transaction data, the storage table association of the standardized enterprise storage data set is carried out to generate time - series associated storage tables, including the bank - enterprise amount association table and the enterprise - customer amount association table. This process helps to systematically manage and associate financial data from different sources, improving data consistency and traceability. The initial check of bills is carried out step by step on the bank - enterprise amount association table and the enterprise - customer amount association table to generate initial check data for step - by - step bill checking. Through the initial check process, potential errors and inconsistencies in the bill data can be found, improving data accuracy. The difference identification of the initial check data for step - by - step bill checking is carried out to generate difference identification data for step - by - step bill checking. This step helps to quickly identify anomalies and differences in the bill data, providing a basis for subsequent data correction. According to the difference identification data for step - by - step bill checking, the bank - enterprise amount association table and the enterprise - customer amount association table are corrected in the ledger to generate a corrected time - series associated storage table. Through this process, errors in the bill data can be corrected to ensure data accuracy and integrity. Through steps such as time - series analysis, storage table association, initial check, and difference identification, problems in the bill data can be systematically discovered and corrected, improving data consistency and accuracy. Through the time - series analysis and step - by - step checking of the bill data, the financial status and transaction patterns of the enterprise can be understood more deeply, enhancing the enterprise's financial management ability and decision - making support.

[0033] Preferably, step S23 includes the following steps:

[0034] Step S231: Perform date matching on the bank - enterprise amount association table and the enterprise - customer amount association table to generate date - matching result data; perform amount matching on the bank - enterprise amount association table and the enterprise - customer amount association table to generate amount - matching result data;

[0035] Step S232: Perform transaction category matching on the bank - enterprise amount association table and the enterprise - customer amount association table according to the date - matching result data and the amount - matching result data to generate transaction category - matching data; integrate the initial check records of the date - matching result data, the amount - matching result data, and the transaction category - matching data to generate initial check data for step - by - step bill checking;

[0036] Step S233: Compare the difference data of the initial check data for step - by - step bill checking to generate difference - comparison data; classify the differences in the initial check data for step - by - step bill checking through the difference - comparison data to generate difference identification data for step - by - step bill checking, where the difference classification includes outstanding item classification, recording error classification, and duplicate record classification.

[0037] The present invention generates matching result data by performing date matching and amount matching on the bank-enterprise amount association table and the enterprise-customer amount association table. This ensures the consistency of transaction records in terms of time and amount, and helps to identify potential errors and anomalies in the data. According to the date matching result data and the amount matching result data, the transaction categories are matched to generate transaction category matching data. By matching the transaction categories, the consistency of transaction records in different dimensions can be further ensured, and the accuracy of data verification can be improved. The date matching result data, the amount matching result data, and the transaction category matching data are integrated to generate preliminary hierarchical bill reconciliation data. This integration process helps to systematically reconcile the bill data, improving the efficiency and accuracy of the reconciliation work. The preliminary hierarchical bill reconciliation data is compared for difference data to generate difference comparison data. Through the difference comparison, the inconsistencies and anomalies in the bill data can be quickly identified, providing a basis for subsequent difference classification and correction. The preliminary hierarchical bill reconciliation data is classified according to the difference comparison data to generate hierarchical bill difference identification data. The difference classification includes outstanding item classification, recording error classification, and duplicate record classification, which helps to take corresponding measures for different types of differences, improving the accuracy and pertinence of data processing. Through detailed date, amount, and transaction category matching, as well as difference comparison and classification, the problems in the bill data can be systematically discovered and corrected, significantly improving the accuracy and consistency of the data. Through refined bill reconciliation and difference identification, enterprises can more comprehensively understand their financial status and transaction patterns, enhancing their financial management and risk control capabilities.

[0038] Preferably, step S3 includes the following steps:

[0039] Step S31: Perform difference review on the time-series association storage correction table to generate review difference result data;

[0040] Step S32: Based on the review difference result data, mark the abnormal bills in the time-series association storage correction table to generate abnormal bill marking data;

[0041] Step S33: Conduct abnormal cause reasoning on the abnormal bill marking data to generate abnormal cause reasoning data; based on the abnormal cause reasoning data, give an early warning of the abnormal cause for the abnormal bill marking data to generate abnormal reconciliation record early warning data;

[0042] Step S34: Verify the abnormal bill marking data through the abnormal reconciliation record early warning data to generate abnormal reconciliation record early warning feedback data.

[0043] The present invention generates review difference result data by performing a difference review on the time-series associated storage correction table. This step ensures that the preliminarily checked and corrected data undergoes further review to confirm the accuracy and completeness of the correction. Based on the review difference result data, abnormal bill marking is performed on the time-series associated storage correction table to generate abnormal bill marking data. By marking abnormal bills, potential problems can be systematically identified and recorded, providing a basis for subsequent processing. Abnormal cause and effect reasoning is performed on the abnormal bill marking data to generate abnormal cause and effect reasoning data. Through cause and effect reasoning, the root causes of abnormal bills can be deeply analyzed, providing more comprehensive abnormal diagnosis information. Abnormal cause early warning is performed on the abnormal bill marking data through the abnormal cause and effect reasoning data to generate abnormal reconciliation record early warning data. This step helps to detect and early warn potential financial problems in advance, improving risk management and prevention capabilities. The abnormal bill marking data is audited and verified through the abnormal reconciliation record early warning data to generate abnormal reconciliation record early warning feedback data. Through audit and verification, the abnormal bill marking can be confirmed and corrected, improving the accuracy and reliability of the data. Through the systematic processing of difference review, abnormal marking, cause and effect reasoning, and early warning feedback, the accuracy and completeness of bill data can be significantly improved, ensuring data quality. Through the marking and cause and effect analysis of abnormal bills, potential financial risks can be timely detected and early warned, enhancing the enterprise's risk management capabilities. The systematic and refined bill review and abnormal handling process helps the enterprise to more efficiently manage and process bill data, improving work efficiency and data processing quality.

[0044] Preferably, step S4 includes the following steps:

[0045] Step S41: Use machine learning methods to generate reconciliation rules for the abnormal reconciliation record early warning feedback data and the hierarchical bill preliminary reconciliation data to obtain an enterprise automatic reconciliation rule engine;

[0046] Step S42: Automatically reconcile the multi-source enterprise bill data based on the enterprise automatic reconciliation rule engine to generate enterprise automatic reconciliation data;

[0047] Step S43: Visualize the enterprise automatic reconciliation data to generate an enterprise automatic reconciliation result report; import the enterprise automatic reconciliation result report into the blockchain for distributed storage of bills to generate enterprise bill block storage data, so as to execute the enterprise data automatic reconciliation storage operation.

[0048] The present invention generates an enterprise automatic reconciliation rule engine by using machine learning methods to generate reconciliation rules for abnormal reconciliation record warning feedback data and hierarchical bill initial reconciliation data. This step utilizes advanced algorithms and data analysis techniques to automatically generate reconciliation rules, improving the efficiency and accuracy of rule generation. Based on the enterprise automatic reconciliation rule engine, automatic reconciliation is performed on multi-source enterprise bill data to generate enterprise automatic reconciliation data. Through automatic reconciliation, the workload of manual reconciliation can be significantly reduced, and the reconciliation efficiency and accuracy can be improved. Data visualization is performed on the enterprise automatic reconciliation data to generate an enterprise automatic reconciliation result report. Data visualization helps to intuitively display the reconciliation results, making it easier for management personnel to understand and analyze the reconciliation data. The enterprise automatic reconciliation result report is imported into the blockchain for distributed storage of bills, thereby generating enterprise bill block storage data. Through blockchain technology, the storage of bill data becomes more secure and transparent, preventing data tampering and loss. By applying the automatic reconciliation rule engine, the reconciliation efficiency can be significantly improved, the reconciliation time can be reduced, and the efficiency of enterprise financial management can be enhanced. By using machine learning and automation technologies, errors in manual reconciliation can be reduced, and the accuracy of reconciliation data can be improved. The application of blockchain technology makes the storage of bill data more secure and reliable, preventing data from being tampered with and lost, and enhancing data security. The systematic and automated reconciliation process helps enterprises optimize the data processing process, improve work efficiency and data processing quality. Through data visualization and the reconciliation result report, management personnel can more intuitively understand the reconciliation situation, enhancing the scientificity and accuracy of management decisions.

[0049] Preferably, step S41 includes the following steps:

[0050] Step S411: Integrate the abnormal reconciliation record warning feedback data and the hierarchical bill initial reconciliation data to generate a comprehensive bill reconciliation data set;

[0051] Step S412: Extract text features from the comprehensive bill reconciliation data set to obtain reconciliation feature data; divide the reconciliation feature data into data sets to generate a model training set and a model test set;

[0052] Step S413: Use the convolutional neural network algorithm to train the model training set to generate an enterprise automatic reconciliation training model; optimize the model hyperparameters of the enterprise automatic reconciliation training engine through the model test set to generate a reconciliation rule generation model;

[0053] Step S414: Generate an enterprise automatic reconciliation rule engine based on the reconciliation rule generation model for the abnormal reconciliation record warning feedback data and the hierarchical bill initial reconciliation data.

[0054] The present invention integrates the early warning feedback data of abnormal reconciliation records and the preliminary reconciliation data of hierarchical bills to generate a comprehensive bill reconciliation dataset. Through data integration, multi-source data can be uniformly processed and analyzed, improving the consistency and integrity of the data. Text feature extraction is performed on the comprehensive bill reconciliation dataset to obtain reconciliation feature data; the reconciliation feature data is partitioned into a dataset to generate a model training set and a model test set. Feature extraction helps to identify and extract key feature data during the reconciliation process, providing high-quality data input for model training. The convolutional neural network algorithm is used to train the model on the model training set to generate an enterprise automatic reconciliation training model. The convolutional neural network algorithm can effectively process and analyze complex reconciliation data, improving the accuracy and prediction ability of the model. The model hyperparameters of the enterprise automatic reconciliation training engine are tuned through the model test set to generate a reconciliation rule generation model. Hyperparameter tuning can optimize the model performance and improve the accuracy and efficiency of reconciliation rule generation. Based on the reconciliation rule generation model, reconciliation rules are generated for the early warning feedback data of abnormal reconciliation records and the preliminary reconciliation data of hierarchical bills to obtain an enterprise automatic reconciliation rule engine. By automatically generating reconciliation rules, the generation efficiency and accuracy of reconciliation rules can be significantly improved. Through automatic reconciliation rule generation, manual intervention can be greatly reduced, the reconciliation efficiency can be improved, and the cost and time of manual reconciliation can be reduced. Using the convolutional neural network algorithm and hyperparameter tuning technology can improve the accuracy of reconciliation rule generation, reduce reconciliation errors, and enhance the reliability of reconciliation data. Using the convolutional neural network algorithm and hyperparameter tuning technology can improve the accuracy of reconciliation rule generation, reduce reconciliation errors, and enhance the reliability of reconciliation data. The systematic and automated reconciliation rule generation process helps enterprises optimize the reconciliation process and improve work efficiency and data processing quality.

[0055] In this specification, an enterprise data automatic reconciliation system is provided for implementing the above-mentioned enterprise data automatic reconciliation method. The enterprise data automatic reconciliation system includes:

[0056] A bill classification module, configured to obtain multi-source enterprise bill data; perform data preprocessing on the multi-source enterprise bill data to generate standard multi-source enterprise bill data; decompose the enterprise bill addresses of the standard multi-source enterprise bill data to generate bill type classification data; and archive and store the bill type classification data in a subordinate database to generate a standardized enterprise storage dataset;

[0057] A preliminary reconciliation module, configured to perform storage table association on the standardized enterprise storage dataset to generate a time-series associated storage table; perform preliminary hierarchical bill reconciliation on the time-series associated storage table to generate preliminary hierarchical bill reconciliation data; and correct the differential ledger of the preliminary hierarchical bill reconciliation data to generate a time-series associated storage correction table;

[0058] An abnormal bill processing module, which is used to conduct a difference review on the time-series associated storage correction table to generate review difference result data; mark abnormal bills on the time-series associated storage correction table based on the review difference result data to generate abnormal bill marking data; give an early warning of the abnormal cause for the abnormal bill marking data to generate abnormal reconciliation record early warning data; and conduct an audit and verification on the abnormal bill marking data through the abnormal reconciliation record early warning data, so as to generate abnormal reconciliation record early warning feedback data.

[0059] An automatic reconciliation rule engine module, which is used to generate reconciliation rules for the abnormal reconciliation record early warning feedback data and the step-by-step bill initial reconciliation data to obtain an enterprise automatic reconciliation rule engine; conduct automatic reconciliation and storage on multi-source enterprise bill data based on the enterprise automatic reconciliation rule engine, so as to generate enterprise bill block storage data to execute the enterprise data automatic reconciliation storage job.

[0060] The beneficial effects of the present invention are as follows: By obtaining enterprise bill data from multiple sources, more comprehensive information can be obtained, which is helpful for subsequent data analysis and processing. By preprocessing the multi-source enterprise bill data, the data can be cleaned and standardized to ensure the accuracy and consistency of the data, providing a reliable basis for subsequent analysis and processing. By performing table association and time-series association storage on the standardized enterprise storage data set, the association relationships between the data can be established, facilitating subsequent bill reconciliation and data query operations. By performing step-by-step initial bill reconciliation on the time-series association storage table, potential bill differences can be identified and marked, and data anomalies can be discovered in advance, which is helpful for timely correction and processing. By performing differential ledger correction and generating a time-series association storage correction table on the step-by-step initial bill reconciliation data, the process of data correction can be recorded and tracked to ensure the accuracy and integrity of the data. By performing differential review on the time-series association storage correction table, the correctness of the data correction can be further confirmed and verified, generating review difference result data, providing a basis for subsequent abnormal bill processing. By performing abnormal bill marking and early warning of abnormal reasons on the time-series association storage correction table, abnormal bills can be quickly discovered and identified, and potential problems can be warned in advance, which is helpful for timely taking measures for processing and adjustment. Through abnormal reconciliation record warning data and audit verification, abnormal bills can be audited and verified to ensure the accuracy and reliability of the abnormal reconciliation records, providing a basis for subsequent processing and feedback. By generating reconciliation rules based on the abnormal reconciliation record warning feedback data and the step-by-step initial bill reconciliation data, an enterprise automatic reconciliation rule engine can be established to realize the automatic reconciliation of multi-source enterprise bill data, improving efficiency and accuracy. By executing the enterprise data automatic reconciliation storage job, the reconciliation results and related data can be stored in the bill block, facilitating subsequent query, analysis and audit operations, and providing data traceability and verifiability. Therefore, the present invention improves the automation level and accuracy of the reconciliation process through systematic data preprocessing, standardized storage, automatic initial verification and correction, intelligent anomaly detection and rule generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a schematic diagram of the step flow of a method for realizing automatic reconciliation of enterprise data;

[0062] Figure 2 is Figure 1 a detailed implementation step flow diagram of step S2 in

[0063] Figure 3 is Figure 1 a detailed implementation step flow diagram of step S3 in

[0064] Figure 4 is Figure 1 a detailed implementation step flow diagram of step S4 in

[0065] The implementation, functional features and advantages of the present invention will be further described in conjunction with the embodiments and with reference to the accompanying drawings. Specific embodiments

[0066] The technical method of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those skilled in the art within the scope of the present invention without creative work belong to the scope of protection of the present invention.

[0067] In addition, the accompanying drawings are only schematic diagrams of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings represent the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0068] It should be understood that although terms such as "first", "second", etc. may be used here to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used here includes any and all combinations of one or more of the listed related items.

[0069] To achieve the above object, please refer to Figures 1 to 4 , a method for realizing automatic reconciliation of enterprise data, the method comprising the following steps:

[0070] Step S1: Obtain multi-source enterprise bill data; perform data preprocessing on the multi-source enterprise bill data to generate standard multi-source enterprise bill data; decompose the enterprise bill addresses of the standard multi-source enterprise bill data to generate bill type classification data; archive and store the bill type classification data in a database to generate a normalized enterprise storage data set;

[0071] Step S2: Perform storage table association on the normalized enterprise storage data set to generate a time-series associated storage table; perform step-by-step initial bill reconciliation on the time-series associated storage table to generate step-by-step initial bill reconciliation data; correct the differential ledger for the step-by-step initial bill reconciliation data to generate a time-series associated storage correction table;

[0072] Step S3: Conduct a difference review on the time-series associated storage correction table to generate review difference result data; perform abnormal bill marking on the time-series associated storage correction table based on the review difference result data to generate abnormal bill marking data; issue an early warning for the cause of the abnormality for the abnormal bill marking data to generate abnormal reconciliation record early warning data; conduct an audit verification on the abnormal bill marking data through the abnormal reconciliation record early warning data, thereby generating abnormal reconciliation record early warning feedback data;

[0073] Step S4: Generate an enterprise automatic reconciliation rule engine by generating reconciliation rules for the abnormal reconciliation record early warning feedback data and the step-by-step bill preliminary reconciliation data; perform automatic reconciliation and storage on the multi-source enterprise bill data based on the enterprise automatic reconciliation rule engine, thereby generating enterprise bill block storage data to execute the enterprise data automatic reconciliation storage operation.

[0074] Through preprocessing and standardization of multi-source enterprise bill data, the present invention ensures that all data formats are unified and generates a standardized enterprise storage data set, which not only improves the accuracy of the data but also makes subsequent bill processing more efficient and consistent. Through step-by-step bill preliminary reconciliation and difference ledger correction, a time-series associated storage correction table is generated, automating the preliminary reconciliation and difference correction processes of the bills, greatly reducing manual operations and human errors, and improving the efficiency and accuracy of reconciliation. Conducting a difference review on the time-series associated storage correction table to generate review difference result data and performing abnormal bill marking and early warning based on these data can promptly detect abnormal bills and generate abnormal reconciliation record early warning data, thereby quickly responding to and handling potential financial problems. By generating an enterprise automatic reconciliation rule engine from the abnormal reconciliation record early warning feedback data and the step-by-step bill preliminary reconciliation data, the reconciliation rules can be dynamically generated and adjusted to ensure that the reconciliation process is more intelligent and personalized, improving the accuracy and adaptability of automatic reconciliation. Storing the reconciled data as enterprise bill block storage data realizes the structured storage and management of the data, improves the retrieval efficiency and storage space utilization rate of the data, and supports the efficient processing and long-term preservation of large-scale enterprise bill data. Therefore, through systematic data preprocessing, standardized storage, automated preliminary verification and correction, intelligent anomaly detection, and rule generation, the present invention improves the automation level and accuracy of the reconciliation process.

[0075] In the embodiment of the present invention, refer to Figure 1 As described, it is a schematic diagram of the step flow of a method for realizing enterprise data automatic reconciliation of the present invention. In this example, the method for realizing enterprise data automatic reconciliation includes the following steps:

[0076] Step S1: Obtain multi-source enterprise bill data; perform data preprocessing on the multi-source enterprise bill data to generate standard multi-source enterprise bill data; decompose the enterprise bill address for the standard multi-source enterprise bill data to generate bill type classification data; archive and store the bill type classification data in a database to generate a standardized enterprise storage dataset;

[0077] In the embodiment of the present invention, bill data is collected from different departments, suppliers, and customer systems of the enterprise, specifically including scanned copies of paper bills, email attachments, CSV or Excel files exported from online systems, etc. Use a data collection tool or API interface to automatically extract bill data from each data source to ensure comprehensive coverage of all bill sources. Remove duplicate data, fill in missing values, and process incorrect data. You can use data cleaning software or write custom cleaning scripts. Convert all bill data into a unified format, such as a unified date format, currency format, and data fields (such as bill number, amount, date, supplier name, etc.). Standardize the bill data from different sources. For example, convert all amounts to the same currency unit and convert dates to a standard format, etc. Use a data conversion tool or programming language (such as the Pandas library in Python) to complete this. Integrate the cleaned and formatted data into a standardized data structure, creating a unified bill database or data table. Verify the integrated data to ensure that all fields and formats meet expectations and the data accuracy is guaranteed. Extract and decompose address information (such as bill address, postal code, city, country) from the bill data. You can use an address resolution service or API (such as the Google Maps API) to automate the process. Associate the extracted address information with the bill data to generate bill address decomposition data, including the detailed address information of each bill. Classify the bills according to the content and characteristics of the bills (such as bill type, purpose, business category). For example, bills can be classified into purchase bills, sales bills, expense bills, etc. Use a data classification algorithm or rule engine (such as keyword-based classification rules, machine learning classification models) to automatically classify the bills. Design a database architecture, create appropriate data tables and fields to store the bill type classification data. Archive and store the bill type classification data by type in different database tables or databases to ensure the structuring of the data and facilitate querying. Use a database management system (such as MySQL, PostgreSQL, MongoDB) for data storage and management, and regularly back up the stored data to prevent data loss or damage. Integrate the archived and stored data into a standardized enterprise storage dataset to ensure that all bill data is stored according to the standard format and classification. Create an index for the stored dataset to optimize query performance and improve data access speed. Perform a final quality check on the standardized dataset to ensure data accuracy, integrity, and consistency.

[0078] Step S2: Perform storage table association on the normalized enterprise storage dataset to generate a time-series associated storage table; perform step-by-step initial bill reconciliation on the time-series associated storage table to generate step-by-step initial bill reconciliation data; perform differential ledger correction on the step-by-step initial bill reconciliation data to generate a time-series associated storage correction table;

[0079] In the embodiment of the present invention, relevant bill data is extracted from the normalized enterprise storage dataset. Determine the fields that need to be time-series associated, such as bill date, transaction number, etc. Sort the bill data in chronological order. The ORDER BY clause in the SQL query statement or the time-series sorting function in the data processing tool can be used. Create a time-series associated storage table to associate data of different bill types in chronological order. The JOIN operation in the database management system can be used to merge relevant data into a new table. Select an appropriate database table structure to support the storage and query of time-series data. Save the associated data to the new table, ensuring that the table can support the query and analysis of time-series data. This table usually contains fields such as bill ID, timestamp, bill amount, transaction type, etc. Develop initial reconciliation rules, such as checking the chronological order of bills, amount consistency, and the integrity of transaction records. Define the reconciliation criteria according to the business logic of the bills. Use a programming language or a database query tool to gradually reconcile the data in the time-series associated storage table. For example, check whether the amount of each transaction matches the corresponding invoice amount, or check whether the bill date meets the expectation. Record each difference or anomaly found during the reconciliation process to generate step-by-step initial bill reconciliation data. This process can be automated through database queries or scripting, and the results can be stored in a new data table. Analyze the differences in the step-by-step initial bill reconciliation data, identify and classify various difference types, such as amount mismatch, bill date error, etc. According to the results of the difference analysis, develop a correction strategy and correct the data in the time-series associated storage table. The correction process can be automatically completed by writing correction scripts or using data processing tools. Save the corrected data to a new table, called the time-series associated storage correction table. This table contains corrected bill records, as well as information such as the reason for correction and the correction time.

[0080] Step S3: Perform differential review on the time-series associated storage correction table to generate review difference result data; perform abnormal bill marking on the time-series associated storage correction table based on the review difference result data to generate abnormal bill marking data; perform early warning of abnormal reasons on the abnormal bill marking data to generate early warning data for abnormal reconciliation records; perform audit verification on the abnormal bill marking data through the early warning data for abnormal reconciliation records, so as to generate early warning feedback data for abnormal reconciliation records;

[0081] In the embodiments of the present invention, by determining the review criteria and rules, such as the bill amount range, time consistency, etc. Set up a review process to ensure that the review rules cover all potential difference types. Review each record in the time-series associated storage correction table. An automated review tool (such as a database query, data analysis software) can be used to check whether the bill records meet the correction criteria. Record the differences and anomalies found during the review process to generate review difference result data. This includes all bill records that fail the review and their difference details. The data can be stored in a new table or file for subsequent processing. According to the difference types and severity levels in the review difference result data, formulate abnormal bill marking rules. For example, mark bills with an amount exceeding a certain range as abnormal. Apply the marking rules to mark the records in the time-series associated storage correction table. A programming script or database update operation can be used to add the marking information to the table, such as adding an "abnormal mark" field. Create a bill data set containing abnormal marks, recording all bill records marked as abnormal and their abnormal types. Save the marked data to a new data table or file. Conduct a detailed analysis of the abnormal bill marking data to identify the causes of anomalies, such as inconsistent bill amounts, duplicate bills, etc. Data analysis tools and algorithms (such as statistical analysis, pattern recognition) can be used to determine the causes of anomalies. Based on the analysis results, generate early warning data for abnormal reconciliation records, specifically including a detailed description of the cause of the anomaly, the early warning level, recommended handling measures, etc. The early warning data can be stored in a dedicated early warning report. Provide the generated early warning data for abnormal reconciliation records to the auditors or automated systems for review. During the review process, verify the accuracy of the cause of the anomaly and check the integrity of the early warning data. According to the review results, adjust or further process the abnormal bill marking data. Generate early warning feedback data for abnormal reconciliation records, including the review results, handling suggestions, and corrective measures. Save the review feedback results to a feedback data table or file for subsequent monitoring and reporting.

[0082] Step S4: Generate reconciliation rules for the early warning feedback data of abnormal reconciliation records and the initial reconciliation data of hierarchical bills to obtain an enterprise automatic reconciliation rule engine; based on the enterprise automatic reconciliation rule engine, automatically reconcile and store multi-source enterprise bill data, thereby generating enterprise bill block storage data to execute the enterprise data automatic reconciliation storage operation.

[0083] In the embodiments of the present invention, by integrating the early warning feedback data of abnormal reconciliation records and the primary reconciliation data of hierarchical bills, a basic data set for rule generation is formed. These data include bill differences, reasons for anomalies, and preliminary reconciliation results. Based on the integrated data, automatic reconciliation rules are defined. These rules may include: Amount matching rule: For example, ensure that the bill amounts match within a preset tolerance range. Date consistency rule: Check whether the dates of the bills conform to business logic. Bill classification rule: Verify the correctness of the bill type and classification. Anomaly handling rule: For bills marked as abnormal, specify the standard operations for handling and correction. Use a rule engine construction tool or programming language (such as Java, Python) to implement the reconciliation rule engine. A rule engine platform (such as Drools, OpenL Tablets) or custom rule engine code can be used to convert the rules into logic that can be automatically executed. Input multi-source enterprise bill data into the enterprise automatic reconciliation rule engine. Data input can be performed through API interfaces, file uploads, or database connections. The rule engine automatically processes the bill data according to the predefined rules. Perform reconciliation operations, including amount comparison, anomaly detection, classification verification, etc. During the automatic reconciliation process, the rule engine will perform a series of reconciliation operations according to the rules, such as: comparing the bill data with the preset standards to check the matching situation; marking and recording all differences or anomalies found during the reconciliation process, generating reconciliation result data, including records of successful reconciliations and the differences found. Store the automatic reconciliation result data in a database or file system. Create an enterprise bill block storage data table or file, containing reconciliation results, difference records, and processing logs. Design the structure of the block storage data, including reconciliation results, bill records, difference explanations, and processing logs. Ensure that the data structure supports efficient querying and storage. Store the reconciliation result data in a database according to the designed data structure, or store it as a blockchain-formatted data file. Data storage can use relational databases (such as MySQL, PostgreSQL) or NoSQL databases (such as MongoDB), or use blockchain technology for distributed storage. Ensure that the automatic reconciliation system can perform storage jobs regularly or on demand to update the enterprise bill block storage data.

[0084] Preferably, step S1 includes the following steps:

[0085] Step S11: Obtain multi-source enterprise bill data;

[0086] Step S12: Perform data preprocessing on the multi-source enterprise bill data to generate standard multi-source enterprise bill data, where data preprocessing includes data cleaning, filling of missing data values, handling of data outliers, and data standardization;

[0087] Step S13: Track the enterprise bill address for the standard multi-source enterprise bill data to generate enterprise bill address tracking data; decompose the standard multi-source enterprise bill data according to the enterprise bill address tracking data to generate bill type classification data, where the bill type classification data includes bank statement data, enterprise financial record data, and customer transaction record data;

[0088] Step S14: Archive and store in the affiliated database based on the bank statement data, enterprise financial record data, and customer transaction record data to generate a standardized enterprise storage data set.

[0089] In the embodiment of the present invention, by identifying and determining each source of enterprise bill data, including bank statements, enterprise financial systems, customer transaction records, etc. Extract bill data from the determined data sources. Data can be collected using methods such as APIs, database queries, file uploads, etc., ensuring that the data formats obtained from different sources are consistent and preparing for subsequent processing. Check and delete duplicate bill records, and correct obvious error data, such as incorrect dates, amounts, etc. Identify the fields and missing patterns of missing data, and fill in the missing values using methods such as mean, median, forward and backward values filling according to the characteristics of the data. Use statistical methods or machine learning models to identify outliers, and correct or delete the outliers according to the actual situation. Unify the data format, for example, standardize the amount to a unified currency unit, and normalize the numerical data to the range of 0-1 or standardize it to a standard distribution with a mean of 0 and a variance of 1. Extract address information from the bill data, standardize the extracted address information, for example, unify the address format and eliminate spelling mistakes. Convert the standardized address information into geographical coordinates (latitude and longitude), and use a geocoding service such as Google Maps API to match the address to the actual location of the enterprise entity to ensure address accuracy. According to the address tracking results, classify the bill data to generate bill type classification data, and divide the bill data into bank statement data, enterprise financial record data, and customer transaction record data. Design a database model for storing enterprise bill data, including table structures and field definitions, and establish appropriate indexes to improve query performance. Import the standardized enterprise storage data set into the database, verify whether the imported data meets the expectations, ensure data integrity and consistency, and formulate a data backup and recovery strategy to prevent data loss or damage.

[0090] Preferably, step S13 includes the following steps:

[0091] Step S131: Parse the multi-source data format of the standard multi-source enterprise bill data to generate multi-source format parsing data;

[0092] Step S132: Extract the bill IP information from the standard multi-source enterprise bill data according to the multi-source format parsing data to obtain multi-source bill IP information data;

[0093] Step S133: Perform external address verification on the multi-source bill IP information data to generate multi-source bill address verification data; perform latitude and longitude coordinate encoding on the multi-source bill IP information data through the multi-source bill address verification data to generate multi-source bill latitude and longitude information data; perform address tracking and matching on the multi-source bill latitude and longitude information data through GIS technology to generate enterprise bill address tracking data;

[0094] Step S134: Decompose the standard multi-source enterprise bill data according to the enterprise bill address tracking data to generate bill type classification data, where the bill type classification data includes bank statement data, enterprise financial record data, and customer transaction record data.

[0095] In the embodiment of the present invention, by identifying the formats of different bill data, including CSV, Excel, JSON, XML, etc. Map bill data in various formats to a unified data model, for example, by defining standard fields (such as date, amount, supplier, etc.). Use data conversion tools (such as ETL tools) to convert data in different formats into a unified format, extract standardized fields, and generate multi-source format parsing data. Verify whether the parsed data meets the expected format and content standards, and correct format errors or data loss problems that occur during the parsing process. Extract IP address information from the bill data. If there is no direct IP field in the bill, it can be inferred or matched through business rules to ensure that the extracted IP information is valid, remove invalid or incorrect IP addresses, and store the extracted IP information as multi-source bill IP information data. Use an external address verification service (such as an IP geolocation service) to verify the multi-source bill IP information data to ensure that the IP address is consistent with the actual geographical location, update the verification result in the multi-source bill IP information data, and generate multi-source bill address verification data. Use a geocoding service (such as Google Maps API, OpenStreetMap) to convert the verified IP information into latitude and longitude coordinates to generate multi-source bill latitude and longitude information data, including latitude and longitude coordinates and their related information. Use GIS technology for address tracking and matching, map the latitude and longitude information to a specific address, and generate enterprise bill address tracking data to ensure the accuracy and integrity of the address match. Classify the bill according to the enterprise bill address tracking data, identify different bill types, and formulate classification rules, such as decomposing according to the bill source (bank, financial record, customer transaction). Store the decomposed bill data according to the type to generate bill type classification data, ensuring that each type of bill data is complete and meets the expectations, including bank statement data, enterprise financial record data, and customer transaction record data. Archive the classified data into a standardized storage system to ensure the queryability and persistence of the data.

[0096] Preferably, step S134 includes the following steps:

[0097] Step S1341: Screen the common address from the enterprise bill address tracking data to obtain the common address bill data and the non-common address bill data;

[0098] Step S1342: Perform the first type of tendency analysis on the standard multi-source enterprise bill data through the common address bill data to generate the first type of tendency data, where the first type of tendency analysis includes bank reconciliation analysis and enterprise financial analysis; perform the second type of tendency analysis on the standard multi-source enterprise bill data through the non-common address bill data to generate the second type of tendency data, where the second type of tendency analysis includes customer transaction record analysis;

[0099] Step S1343: Extract the bill content features from the standard multi-source enterprise bill data to obtain the bill content feature data; decompose the bill content feature data according to the first type of tendency data and the second type of tendency data to generate the bill type classification data, where the bill type classification data includes bank statement data, enterprise financial record data, and customer transaction record data.

[0100] In the embodiments of the present invention, by determining what is a "frequently used address", for example, an address that appears frequently or an address related to a specific business, screening rules are set, such as the frequency of address appearance or importance, etc. The appearance frequency of each address in the enterprise bill address tracking data is counted. Based on the frequency or other criteria, the frequently used address bill data is screened out. The unselected addresses are marked as non-frequently used addresses, and the non-frequently used address bill data is generated. Bank statement data is extracted from the frequently used address bill data, and the bank statement data is analyzed using data analysis methods (such as time series analysis, reconciliation difference analysis). Enterprise financial record data is extracted, and financial trend analysis, financial ratio analysis, etc. are performed on the enterprise financial record data. The results of bank reconciliation analysis and enterprise financial analysis are combined to generate the first type of tendency data. Customer transaction records in the non-frequently used address bill data are extracted, and behavior pattern analysis, transaction frequency analysis, etc. are performed on the customer transaction records. The second type of tendency data is generated according to the analysis results of the customer transaction records. Key features in the bill content, such as amount, date, transaction party, etc., are identified, and bill content feature data is extracted using natural language processing (NLP) technology or a rule engine. The extracted feature data is sorted out to form a bill content feature data set. Using statistical analysis or machine learning algorithms, the first type of tendency data is combined with the bill content feature data, and the bill content feature data is decomposed into bank statement data and enterprise financial record data. Analyzing in combination with the second type of tendency data, the bill content feature data is decomposed into customer transaction record data. The first type of tendency data and the second type of tendency data are comprehensively used to finally classify the bill, and bill type classification data including bank statement data, enterprise financial record data, and customer transaction record data is generated. According to the structure of the bill type classification data, it is stored in a normalized database, and the classified bill data is backed up regularly to prevent data loss. The accuracy of data classification is checked regularly, and necessary updates and corrections are made.

[0101] As an example of the present invention, refer to Figure 2 shown, in this example, step S2 includes:

[0102] Step S21: Perform time-series amount transaction analysis on the normalized enterprise storage data set to generate time-series amount transaction data;

[0103] Step S22: Perform storage table association on the normalized enterprise storage data set through the time-series amount transaction data to generate a time-series associated storage table, where the time-series associated storage table includes a bank-enterprise amount association table and an enterprise-customer amount association table;

[0104] Step S23: Perform step-by-step bill preliminary reconciliation on the bank-enterprise amount association table and the enterprise-customer amount association table to generate step-by-step bill preliminary reconciliation data; perform difference identification on the step-by-step bill preliminary reconciliation data to generate step-by-step bill difference identification data;

[0105] Step S24: Amend the bank-enterprise amount association table and the enterprise-customer amount association table according to the hierarchical bill difference identification data, and generate a time-series association storage amendment table.

[0106] In the embodiment of the present invention, relevant transaction data, including amount, transaction date, trading party, etc., are extracted from the normalized enterprise storage dataset, and missing values, duplicate data, and abnormal data are processed to ensure data quality. Determine the time granularity (such as daily, weekly, monthly), and sort and group the transaction data by time. Use time series analysis methods (such as moving average, exponential smoothing) to analyze the change trend of the amount transactions. Create a time-series amount transaction dataset to record the transaction amount, trend, and pattern within each time period. Define how to associate the data in different tables. For example, through the common fields (such as account number, transaction amount) in the bank-enterprise transaction data and the enterprise-customer transaction data. Extract the transaction records between the bank and the enterprise from the time-series amount transaction data. Create a bank-enterprise amount association table, which includes the transaction amount and its time-series data between the bank account and the enterprise account. Extract the transaction records between the enterprise and the customer from the time-series amount transaction data. Create an enterprise-customer amount association table, which includes the transaction amount and its time-series data between the enterprise account and the customer account. Extract the data that needs to be checked from the bank-enterprise amount association table and the enterprise-customer amount association table. Conduct the check level by level. For example, compare the amount, time, and counterparty account information of each transaction to ensure the consistency and accuracy of the records. Record the findings during the check process to generate the initial hierarchical bill check data. Compare the initial hierarchical bill check data to identify differences such as inconsistent amounts, missing or incorrect transaction records. Mark and record the difference data to generate hierarchical bill difference identification data, including the detailed information of the differences and possible reasons. Develop an amendment plan. Based on the hierarchical bill difference identification data, determine the ledger entries that need to be modified and the amendment methods. Amend the bank-enterprise amount association table and the enterprise-customer amount association table, and update the ledger records according to the difference identification data. Create a time-series association storage amendment table to record the amended data, including the comparison information before and after the amendment and the details of the amendment operations. Confirm whether the amended data is correct, conduct the check and verification again, and record the details of all amendment operations to ensure the transparency and traceability of the ledger.

[0107] Preferably, step S23 includes the following steps:

[0108] Step S231: Perform date matching on the bank-enterprise amount association table and the enterprise-customer amount association table to generate date matching result data; perform amount matching on the bank-enterprise amount association table and the enterprise-customer amount association table to generate amount matching result data;

[0109] Step S232: Perform transaction category matching on the bank-enterprise amount association table and the enterprise-customer amount association table according to the date matching result data and the amount matching result data to generate transaction category matching data; integrate the date matching result data, the amount matching result data, and the transaction category matching data for preliminary verification record to generate step-by-step bill preliminary verification data;

[0110] Step S233: Compare the difference data for the step-by-step bill preliminary verification data to generate difference comparison data; classify the step-by-step bill preliminary verification data through the difference comparison data to generate step-by-step bill difference identification data, where the difference classification includes outstanding item classification, recording error classification, and duplicate record classification.

[0111] In the embodiments of the present invention, the date field of the transaction records is extracted from the bank-enterprise amount association table and the enterprise-customer amount association table. A date matching rule is set, for example, transactions on the same date are matched, and a certain time window (such as 24 hours) is allowed to cope with time zone differences. The dates in the bank-enterprise amount association table are matched with the dates in the enterprise-customer amount association table, and the transactions with successful and failed matches are recorded to generate date matching result data. The amount fields are extracted from the bank-enterprise amount association table and the enterprise-customer amount association table. An amount matching rule is set, for example, the amounts are directly matched or a certain error range (such as 5%) is allowed. The amounts in the bank-enterprise amount association table are matched with the amounts in the enterprise-customer amount association table, and the transactions with successful and failed matches are recorded to generate amount matching result data. The transaction categories are defined, such as income, expenditure, expenses, etc., and are matched according to the transaction category field (if it exists). If the transaction category field is missing, the transaction description or other fields need to be used for speculation. The transaction categories in the bank-enterprise amount association table and the enterprise-customer amount association table are matched, and the categories with successful and failed matches are recorded to generate transaction category matching data. The date matching result data, the amount matching result data, and the transaction category matching data are combined to form a comprehensive preliminary verification record, and a step-by-step bill preliminary verification data is created, including the matching results and the integrated records. The data that needs to be compared is extracted from the step-by-step bill preliminary verification data, the differences in the matching results are compared, problems such as unmatched, inconsistent amounts, and inconsistent categories are identified, all detected differences are recorded, and difference comparison data is generated. A difference classification standard is formulated to identify those account items that appear in the bank-enterprise amount association table or the enterprise-customer amount association table but do not appear in the other table, identify the account items with incorrect amount, date, or category records, identify those account items that are repeatedly recorded in the two tables, mark each difference record according to the classification rules, organize the marked data, and generate step-by-step bill difference identification data. The step-by-step bill preliminary verification data, the difference comparison data, and the step-by-step bill difference identification data are stored in the database, these data are regularly backed up to prevent loss, and the accuracy of data classification and identification is regularly checked to ensure the integrity of difference classification and verification records.

[0112] As an example of the present invention, refer to Figure 3 shown, in this example, step S3 includes:

[0113] Step S31: Perform a difference review on the time-series association storage correction table to generate review difference result data;

[0114] Step S32: Based on the review difference result data, mark the abnormal bills in the time-series association storage correction table to generate abnormal bill marking data;

[0115] Step S33: Perform abnormal cause and effect reasoning on the abnormal bill marking data to generate abnormal cause and effect reasoning data; use the abnormal cause and effect reasoning data to give early warnings of abnormal causes for the abnormal bill marking data to generate abnormal reconciliation record early warning data;

[0116] Step S34: Use the abnormal reconciliation record early warning data to conduct audit verification on the abnormal bill marking data, thereby generating abnormal reconciliation record early warning feedback data.

[0117] In the embodiment of the present invention, by extracting the records to be reviewed from the time-series associated storage correction table, including the corrected data and the original difference records. Set review rules, such as comparing fields such as the corrected amount, date, and category to ensure the accuracy of the correction operation. Compare the corrected data with the differences in the initial data one by one to identify whether there are any omissions or incorrect corrections. Record any differences found during the review process to generate review difference result data. Extract information from the review difference result data generated in step S31. Define abnormal marking criteria, such as the amount deviation exceeding the threshold, frequent corrections, abnormal transaction patterns, etc. Analyze the review difference result data according to the marking rules to identify abnormal bills and generate abnormal bill marking data to mark the abnormal bills that need further analysis. Extract the abnormal records to be analyzed from the abnormal bill marking data, apply causal reasoning techniques (such as Bayesian networks, causal diagram models) to analyze the causes of abnormal bills, identify potential causal relationships, generate abnormal cause and effect reasoning data, and record the causes and associated factors of each abnormal bill. Set early warning rules according to the abnormal cause and effect reasoning data, such as generating early warning signals according to the identified abnormal causes, generating abnormal reconciliation record early warning data, and recording early warning information and further measures required. Extract information from the abnormal reconciliation record early warning data generated in step S33, define audit criteria, such as whether further investigation is required, whether the early warning conditions are met, etc. Review the abnormal reconciliation record early warning data to confirm whether the early warning is reasonable, conduct investigations as needed, record the audit results and any additional problems found, and generate abnormal reconciliation record early warning feedback data. Store the abnormal reconciliation record early warning feedback data in the database for subsequent reference and processing, and take appropriate actions according to the audit results, such as conducting a detailed investigation, correcting the bill, or updating internal control measures.

[0118] As an example of the present invention, refer to Figure 4 shown, in this example, step S4 includes:

[0119] Step S41: Use machine learning methods to generate reconciliation rules for the abnormal reconciliation record early warning feedback data and the step-by-step bill initial reconciliation data to obtain an enterprise automatic reconciliation rule engine;

[0120] Step S42: Automatically reconcile the multi-source enterprise bill data based on the enterprise automatic reconciliation rule engine to generate enterprise automatic reconciliation data;

[0121] Step S43: Visualize the enterprise automatic reconciliation data to generate an enterprise automatic reconciliation result report; import the enterprise automatic reconciliation result report into the blockchain for distributed storage of bills, thereby generating enterprise bill block storage data to execute the enterprise data automatic reconciliation storage job.

[0122] In the embodiment of the present invention, relevant information is extracted from the exception reconciliation record warning feedback data generated in step S33 and the step-by-step bill preliminary reconciliation data generated in step S23. Ensure that the data formats are consistent, handle missing values and outliers to facilitate the training of the machine learning model. Extract features from the warning feedback data and the preliminary reconciliation data, such as transaction amount, date, transaction category, matching result, etc. Select the features that are most useful for generating reconciliation rules, such as bill amount deviation, matching category, etc. Select a machine learning algorithm suitable for generating reconciliation rules, such as decision tree, random forest, support vector machine (SVM), or neural network. Use the preprocessed data to train the machine learning model. Set up training sets and test sets, and use cross-validation to optimize the model parameters. Use the trained model to predict the data and generate reconciliation rules. For example, according to the decision rules output by the model, generate specific automatic reconciliation strategies and rules. Store the generated reconciliation rules in the rule engine to build an enterprise automatic reconciliation rule engine. Extract the bill records to be reconciled from the multi-source enterprise bill data, ensuring that the bill data formats are consistent and meet the requirements of the rule engine. Apply the enterprise automatic reconciliation rule engine to the multi-source enterprise bill data. Use the generated rules to match and compare the bill data. Automatically perform reconciliation operations, including amount matching, date matching, transaction category matching, etc., to generate enterprise automatic reconciliation data, including records of successful and failed matches. Select an appropriate data visualization tool (such as Tableau, PowerBI, D3.js, etc.) to display the enterprise automatic reconciliation results, create visualization charts and reports, including an overview of the reconciliation results, statistics of successful and failed records, distribution of abnormal bills, etc., to generate an enterprise automatic reconciliation result report containing detailed reconciliation analysis and summary. Select a suitable blockchain platform (such as Ethereum, Hyperledger Fabric, etc.) for distributed storage of bill data, convert the enterprise automatic reconciliation result report into a blockchain-storable data format (such as JSON or XML), import the converted data into the blockchain, create blockchain transaction records to store the reconciliation results, verify the integrity and security of the data, and ensure that the distributed storage of data on the blockchain meets expectations. Monitor the execution of the automatic reconciliation storage job to ensure that the data is successfully stored on the blockchain, and regularly maintain and update the data stored on the blockchain to keep the bill data accurate and up-to-date.

[0123] Preferably, step S41 includes the following steps:

[0124] Step S411: Integrate the early warning feedback data of abnormal reconciliation records and the preliminary reconciliation data of hierarchical bills to generate a comprehensive bill reconciliation dataset;

[0125] Step S412: Extract text features from the comprehensive bill reconciliation dataset to obtain reconciliation feature data; Divide the reconciliation feature data into datasets to generate a model training set and a model test set;

[0126] Step S413: Use the convolutional neural network algorithm to train the model training set to generate an enterprise automatic reconciliation training model; Optimize the model hyperparameters of the enterprise automatic reconciliation training engine through the model test set to generate a reconciliation rule generation model;

[0127] Step S414: Generate reconciliation rules for the early warning feedback data of abnormal reconciliation records and the preliminary reconciliation data of hierarchical bills based on the reconciliation rule generation model to obtain an enterprise automatic reconciliation rule engine.

[0128] In the embodiments of the present invention, relevant information is extracted from the abnormal reconciliation record warning feedback data generated in step S33 and the hierarchical bill preliminary reconciliation data generated in step S23, and the missing values and outliers in the data are processed to ensure data quality. The abnormal reconciliation record warning feedback data and the hierarchical bill preliminary reconciliation data are merged to form a comprehensive bill reconciliation data set, ensuring that the integrated data format is consistent and suitable for subsequent analysis and processing. Important features related to reconciliation are identified, such as transaction amount, date, transaction category, reconciliation status, etc. For text-type data, natural language processing techniques (such as TF-IDF, word embedding) are used to extract features to generate reconciliation feature data, which contains all the extracted features. The reconciliation feature data is divided into a model training set and a model test set. Usually, 70%-80% of the data is used for training, and the remaining data is used for testing. If the data set is imbalanced, data balancing processing (such as oversampling or undersampling) is performed. A convolutional neural network architecture is designed, including convolutional layers, pooling layers, and fully connected layers, and the model training set is used for training. The loss function is optimized and the model weights are adjusted, and training parameters are set, such as learning rate, batch size, and number of training epochs. Hyperparameters to be tuned are selected, such as convolutional kernel size, number of layers, learning rate, etc., and the model is hyperparameter-tuned using grid search or random search methods to select the best combination of hyperparameters. The performance of the model is verified through the model test set to ensure the effect of hyperparameter tuning. Metrics such as accuracy, recall rate, and F1 score of the model are evaluated, and the model is further optimized according to the evaluation results to ensure that the generated rules are effective and reliable. The reconciliation rule generation model is used to generate rules for the abnormal reconciliation record warning feedback data and the hierarchical bill preliminary reconciliation data, and the generated reconciliation rules are extracted to determine which rules are applicable to the automatic reconciliation process. The generated reconciliation rules are integrated into the enterprise automatic reconciliation rule engine, and a software system for the reconciliation rule engine is developed to ensure that it can apply the generated rules in real time for automatic reconciliation and verify the effect of the rule engine to ensure that it can effectively handle different types of bill reconciliation situations.

[0129] In this specification, an enterprise data automatic reconciliation system is provided for implementing the above-mentioned method for enterprise data automatic reconciliation. The enterprise data automatic reconciliation system includes:

[0130] A bill classification module for obtaining multi-source enterprise bill data; performing data preprocessing on the multi-source enterprise bill data to generate standard multi-source enterprise bill data; decomposing the enterprise bill addresses of the standard multi-source enterprise bill data to generate bill type classification data; and archiving and storing the bill type classification data in a subordinated database to generate a standardized enterprise storage data set.

[0131] The initial verification module is used to perform storage table association on the standardized enterprise storage data set to generate a time-series associated storage table; perform step-by-step bill initial verification on the time-series associated storage table to generate step-by-step bill initial verification data; perform differential ledger correction on the step-by-step bill initial verification data to generate a time-series associated storage correction table;

[0132] The abnormal bill processing module is used to perform differential review on the time-series associated storage correction table to generate review difference result data; mark abnormal bills on the time-series associated storage correction table based on the review difference result data to generate abnormal bill mark data; give early warnings about the reasons for abnormalities for the abnormal bill mark data to generate abnormal reconciliation record early warning data; perform audit verification on the abnormal bill mark data through the abnormal reconciliation record early warning data, thereby generating abnormal reconciliation record early warning feedback data;

[0133] The automatic reconciliation rule engine module is used to generate reconciliation rules for the abnormal reconciliation record early warning feedback data and the step-by-step bill initial verification data to obtain an enterprise automatic reconciliation rule engine; perform automatic reconciliation and storage on the multi-source enterprise bill data based on the enterprise automatic reconciliation rule engine, thereby generating enterprise bill block storage data to execute the enterprise data automatic reconciliation storage job.

[0134] The beneficial effects of the present invention are as follows: By obtaining enterprise bill data from multiple sources, more comprehensive information can be obtained, which is helpful for subsequent data analysis and processing. By preprocessing multi-source enterprise bill data, the data can be cleaned and standardized to ensure the accuracy and consistency of the data, providing a reliable basis for subsequent analysis and processing. By performing table association and time-series association storage on the standardized enterprise storage data set, the association relationships between data can be established, facilitating subsequent bill reconciliation and data query operations. By performing step-by-step initial bill reconciliation on the time-series association storage table, potential bill differences can be identified and marked, and data anomalies can be detected in advance, which is helpful for timely correction and processing. By performing difference ledger correction and generating a time-series association storage correction table on the step-by-step initial bill reconciliation data, the process of data correction can be recorded and tracked to ensure the accuracy and integrity of the data. By performing difference review on the time-series association storage correction table, the correctness of data correction can be further confirmed and verified, generating review difference result data, providing a basis for subsequent abnormal bill processing. By performing abnormal bill marking and abnormal cause warning on the time-series association storage correction table, abnormal bills can be quickly discovered and identified, and potential problems can be warned in advance, which is helpful for timely taking measures for processing and adjustment. By auditing and verifying the abnormal reconciliation record warning data, the abnormal bills can be audited and verified to ensure the accuracy and reliability of the abnormal reconciliation records, providing a basis for subsequent processing and feedback. By generating reconciliation rules based on the abnormal reconciliation record warning feedback data and the step-by-step initial bill reconciliation data, an enterprise automatic reconciliation rule engine can be established to achieve automatic reconciliation of multi-source enterprise bill data, improving efficiency and accuracy. By executing the enterprise data automatic reconciliation storage job, the reconciliation results and relevant data can be stored in the bill block, facilitating subsequent query, analysis, and auditing operations, and providing data traceability and verifiability. Therefore, the present invention improves the automation level and accuracy of the reconciliation process through systematic data preprocessing, standardized storage, automated initial verification and correction, intelligent anomaly detection, and rule generation.

[0135] Therefore, in any aspect, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, it is intended to embrace all changes within the meaning and scope of the equivalent elements of the application documents in the present invention.

[0136] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for realizing automatic reconciliation of enterprise data, characterized in that: The following steps are involved: Step S1: Obtain multi-source enterprise billing data; Perform data preprocessing on multi-source enterprise billing data to generate standard multi-source enterprise billing data; Decompose the corporate billing address of standard multi-source corporate billing data to generate billing type classification data; The bill type classification data is archived and stored in a separate database to generate a standardized enterprise storage data set; step S1 includes the following steps: Step S11: Acquire multi-source enterprise billing data; Step S12: preprocessing the multi-source enterprise billing data to generate standard multi-source enterprise billing data, wherein the data preprocessing includes data cleaning, data missing value filling, data outlier processing and data standardization; Step S13: tracking the corporate billing address of the standard multi-source corporate billing data to generate corporate billing address tracking data; performing bill decomposition on the standard multi-source corporate billing data according to the corporate billing address tracking data to generate bill type classification data, wherein the bill type classification data includes bank statement data, corporate financial record data, and customer transaction record data; Step S13 includes the following steps: Step S131: parsing the standard multi-source enterprise billing data in a multi-source data format to generate multi-source format parsed data; Step S132: extracting billing IP information from standard multi-source enterprise billing data according to the multi-source format parsing data to obtain multi-source billing IP information data; Step S133: Perform external address verification on the multi-source billing IP information data to generate multi-source billing address verification data; perform latitude and longitude coordinate encoding on the multi-source billing IP information data through the multi-source billing address verification data to generate multi-source billing latitude and longitude information data; perform address tracking and matching on the multi-source billing latitude and longitude information data through GIS technology to generate corporate billing address tracking data; Step S134: Decomposing the standard multi-source enterprise billing data according to the enterprise billing address tracking data to generate billing type classification data, wherein the billing type classification data includes bank statement data, enterprise financial record data, and customer transaction record data; Step S134 includes the following steps: Step S1341: screening the enterprise billing address tracking data for common addresses to obtain common address billing data and uncommon address billing data; Step S1342: Perform bank reconciliation analysis and enterprise financial analysis on the standard multi-source enterprise billing data through the commonly used address billing data to generate the first type of trend data; perform customer transaction record analysis on the standard multi-source enterprise billing data through the uncommon address billing data to generate the second type of trend data; Step S1343: extracting bill content features from the standard multi-source enterprise bill data to obtain bill content feature data; performing bill decomposition on the bill content feature data according to the first type tendency data and the second type tendency data to generate bill type classification data, wherein the bill type classification data includes bank statement data, enterprise financial record data, and customer transaction record data; Step S14: archiving and storing the bank statement data, enterprise financial record data and customer transaction record data in separate databases to generate a standardized enterprise storage data set; Step S2: associate the storage tables of the standardized enterprise storage data set to generate a time-series associated storage table; perform a level-by-level initial bill check on the time-series associated storage table to generate level-by-level initial bill check data; perform a difference account book correction on the level-by-level initial bill check data to generate a time-series associated storage correction table; Step S3: review the difference of the time series associated storage correction table to generate review difference result data; mark the time series associated storage correction table with abnormal bills based on the review difference result data to generate abnormal bill mark data; warn the abnormal cause of the abnormal bill mark data to generate abnormal reconciliation record warning data; review and verify the abnormal bill mark data through the abnormal reconciliation record warning data, thereby generating abnormal reconciliation record warning feedback data; Step S4: Generate reconciliation rules for abnormal reconciliation record warning feedback data and level-by-level bill initial verification data to obtain an enterprise automatic reconciliation rule engine; automatically reconcile and store multi-source enterprise bill data based on the enterprise automatic reconciliation rule engine, thereby generating enterprise bill block storage data to execute enterprise data automatic reconciliation storage operations.

2. The method for realizing automatic reconciliation of enterprise data according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: Perform time series amount transaction analysis on the standardized enterprise storage data set to generate time series amount transaction data; Step S22: Associating the storage tables of the normalized enterprise storage data set with the time series amount transaction data to generate a time series associated storage table, wherein the time series associated storage table includes a bank-enterprise amount association table and an enterprise-customer amount association table; Step S23: Performing a level-by-level initial bill check on the bank-enterprise amount association table and the enterprise-customer amount association table to generate level-by-level initial bill check data; performing difference identification on the level-by-level initial bill check data to generate level-by-level bill difference identification data; Step S24: amend the bank-enterprise amount association table and the enterprise-customer amount association table according to the level-by-level bill difference identification data, and generate a time-series association storage amendment table.

3. The method for realizing automatic reconciliation of enterprise data according to claim 2, characterized in that: Step S23 includes the following steps: Step S231: performing date matching on the bank-enterprise amount association table and the enterprise-customer amount association table to generate date matching result data; performing amount matching on the bank-enterprise amount association table and the enterprise-customer amount association table to generate amount matching result data; Step S232: performing transaction category matching on the bank-enterprise amount association table and the enterprise-customer amount association table according to the date matching result data and the amount matching result data, and generating transaction category matching data; performing preliminary verification and record integration on the date matching result data, the amount matching result data and the transaction category matching data, and generating preliminary verification data of the level-by-level bills; Step S233: Perform difference data comparison on the initial verification data of the level-by-level bills to generate difference comparison data; perform difference classification on the initial verification data of the level-by-level bills through the difference comparison data to generate level-by-level bill difference identification data, wherein the difference classification includes classification of uncollected items, classification of record errors and classification of duplicate records.

4. The method for realizing automatic reconciliation of enterprise data according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: review the difference of the time series associated storage correction table to generate review difference result data; Step S32: marking the abnormal bill in the time series associated storage correction table based on the review difference result data, and generating abnormal bill marking data; Step S33: performing abnormal causal reasoning on the abnormal bill mark data to generate abnormal causal reasoning data; performing abnormal cause warning on the abnormal bill mark data through the abnormal causal reasoning data to generate abnormal reconciliation record warning data; Step S34: The abnormal billing mark data is audited and verified through the abnormal reconciliation record warning data, thereby generating abnormal reconciliation record warning feedback data.

5. The method for realizing automatic reconciliation of enterprise data according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: Generate reconciliation rules for abnormal reconciliation record warning feedback data and level-by-level bill initial verification data using a machine learning method to obtain an enterprise automatic reconciliation rule engine; Step S42: Automatically reconcile multi-source enterprise billing data based on the enterprise automatic reconciliation rule engine to generate enterprise automatic reconciliation data; Step S43: Visualize the enterprise automatic reconciliation data to generate an enterprise automatic reconciliation result report; import the enterprise automatic reconciliation result report into the blockchain for distributed bill storage, thereby generating enterprise bill block storage data to perform enterprise data automatic reconciliation storage operations.

6. The method for realizing automatic reconciliation of enterprise data according to claim 5, characterized in that: Step S41 includes the following steps: Step S411: Integrate the abnormal reconciliation record warning feedback data and the level-by-level bill preliminary verification data to generate a comprehensive bill verification data set; Step S412: extract text features from the comprehensive data set of the reconciliation statement to obtain reconciliation feature data; divide the reconciliation feature data into data sets to generate a model training set and a model test set; Step S413: Perform model training on the model training set using a convolutional neural network algorithm to generate an enterprise automatic account reconciliation training model; perform model hyperparameter tuning on the enterprise automatic account reconciliation training model using a model test set to generate an account reconciliation rule generation model; Step S414: Based on the reconciliation rule generation model, reconciliation rules are generated for the abnormal reconciliation record warning feedback data and the level-by-level bill initial verification data to obtain the enterprise automatic reconciliation rule engine.

7. An automatic reconciliation system for enterprise data, characterized in that: Used to execute the method for realizing automatic reconciliation of enterprise data as claimed in claim 1, the system for realizing automatic reconciliation of enterprise data comprises: The bill classification module is used to obtain multi-source enterprise bill data; perform data preprocessing on the multi-source enterprise bill data to generate standard multi-source enterprise bill data; perform enterprise bill address decomposition on the standard multi-source enterprise bill data to generate bill type classification data; perform database archiving and storage on the bill type classification data to generate a standardized enterprise storage data set; The initial verification module is used to associate the storage tables of the standardized enterprise storage data set to generate a time-series associated storage table; to perform a level-by-level initial verification of the time-series associated storage table to generate level-by-level initial verification data of the bills; to perform a difference account book correction on the level-by-level initial verification data to generate a time-series associated storage correction table; The abnormal bill processing module is used to review the difference of the time series associated storage correction table and generate the review difference result data; based on the review difference result data, the time series associated storage correction table is marked with abnormal bills to generate abnormal bill marking data; the abnormal cause warning is issued to the abnormal bill marking data to generate abnormal reconciliation record warning data; the abnormal bill marking data is audited and verified through the abnormal reconciliation record warning data, thereby generating abnormal reconciliation record warning feedback data; The automatic reconciliation rule engine module is used to generate reconciliation rules for abnormal reconciliation record warning feedback data and step-by-step bill initial verification data to obtain the enterprise automatic reconciliation rule engine; based on the enterprise automatic reconciliation rule engine, multi-source enterprise billing data is automatically reconciled and stored, thereby generating enterprise billing block storage data to execute enterprise data automatic reconciliation and storage operations.

Images