A non-tax special account supervision system and method based on electronic accounting vouchers

Abstract

The present invention discloses a non-tax special account supervision system and method based on electronic accounting vouchers, which relates to the field of data security supervision technology, including a supervision center, wherein the supervision center is communicatively connected with a data integration module, a data storage module, a data integration module, an anomaly detection module, an anomaly handling response module and a report analysis module, wherein the modules are connected by electrical signals. The present invention significantly improves the supervision efficiency through automated data integration and anomaly detection mechanisms, can monitor and analyze a large amount of electronic accounting voucher data in real time, and quickly identify abnormal situations. Compared with manual audit methods, the automated process reduces the possibility of human errors and omissions, and also reduces the work pressure of supervisors. In addition, a built-in risk warning mechanism immediately triggers the warning process once an abnormality or potential risk is found, which helps to take measures quickly to prevent the risk from expanding and spreading.

Description

Technical Field

[0001] The present invention relates to the technical field of data security supervision, and in particular to a non-tax special account supervision system and method based on electronic accounting vouchers. Background Art

[0002] The supervision of non-tax special accounts is an important part of the fiscal management system. It mainly involves the standardized management and supervision of non-tax revenue to ensure the rational use and transparent supervision of these funds. With the continuous advancement and development of information technology, the application of electronic accounting vouchers provides a technical foundation for the automation and intelligence of the non-tax special account supervision system.

[0003] For example, a method for dynamic supervision of fiscal and tax revenue and expenditure data security with Chinese patent publication number: CN112837139A involves the technical field of data security supervision, and the method includes: S1, obtaining a communication path; S2, establishing multiple data transfer nodes; S3, obtaining any three adjacent data transfer nodes as a first data monitoring node, a second data monitoring node and a third data monitoring node; S4, controlling the first data monitoring node to receive a source data set, controlling the first data monitoring node to store the source data set to form a first temporary database, and controlling the first data monitoring node to send the source data set to the second data monitoring node; S5, controlling the second data monitoring node to receive the source data set sent by the first data monitoring node, and determining whether the received source data set is complete.

[0004] In the prior art, real-time dynamic monitoring is carried out during the transmission process of fiscal revenue and expenditure data, abnormal situations are reminded, and new data communication paths are established by hiding abnormal nodes for response processing. However, the non-tax special account supervision system involves data integration among multiple systems, and there are problems with data compatibility and inconsistent abnormal identification standards. Therefore, how to improve data compatibility, divide unified abnormal identification standards, and evaluate and warn of the severity of abnormal situations are the problems we need to solve. To this end, a non-tax special account supervision system and method based on electronic accounting vouchers are proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a non-tax special account supervision system and method based on electronic accounting vouchers to solve the problems raised in the above-mentioned background technology.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] In a first aspect, a non-tax special account supervision system based on electronic accounting vouchers includes a supervision center, wherein the supervision center is communicatively connected to a data integration module, a data storage module, a data integration module, an anomaly detection module, an anomaly processing response module, and a report analysis module, wherein the modules are connected by electrical signals;

[0008] The data integration module is used to collect electronic accounting voucher data related to non-tax revenue from various business systems, including non-tax revenue collection system, integrated budget management system, debt management system, etc., to achieve unified access of multi-source data, provide a basis for subsequent processing, and perform format conversion and standardization processing on the collected electronic accounting voucher data to ensure data consistency and compatibility, providing a basis for subsequent data processing and analysis;

[0009] The data storage module uses a high-performance, high-reliability data warehouse to store processed electronic accounting voucher data, ensures data security and integrity, and provides data backup, recovery, and query functions to support rapid retrieval and access to data, providing data support for regulatory work;

[0010] The data integration module, through data integration technology, merges, removes duplication and associates data from different business systems to form a unified data view;

[0011] The anomaly detection module presets anomaly identification standards, analyzes and detects electronic accounting voucher data, identifies abnormal data, and further analyzes and judges the detected abnormal data to confirm whether it is a real abnormal situation and gives a specific abnormal type;

[0012] The exception handling response module triggers corresponding processing procedures for detected abnormal situations, such as notifying relevant personnel, freezing relevant accounts, etc., dynamically adjusts the data transmission path, bypasses abnormal nodes, ensures the continuity and stability of data transmission, and records the exception handling process and results;

[0013] The report analysis module is used to conduct in-depth mining and analysis of electronic accounting voucher data, generate various regulatory reports and statistical reports based on the analysis results, and provide data support.

[0014] A further improvement of the technical solution of the present invention is that in the data integration module, the process of collecting electronic accounting voucher data related to non-tax revenue is:

[0015] Identify all business systems that need to be integrated, including but not limited to non-tax revenue collection system, budget management integrated system, debt management system, etc. Establish data access protocols for each business system and access the electronic accounting voucher data of each business system through API interface calls;

[0016] Specify the data types and fields of electronic accounting vouchers to be extracted from each business system, namely invoice number, invoice date, amount, payee, and payee information;

[0017] For each business system, configure the corresponding interface docking program, such as using Web Service, REST API or direct database connection, and use ETL tools to extract electronic accounting voucher data related to non-tax revenue from each business system at predetermined time intervals;

[0018] Clean the acquired electronic accounting voucher data related to non-tax revenue, remove duplicates, errors and incomplete records, and conduct preliminary data verification during the data access process to ensure the accuracy and completeness of the data;

[0019] Establish data mapping relationships, convert data formats of different business systems into a unified format, set up monitoring mechanisms to monitor any anomalies in the data integration process, and record detailed operation logs to facilitate problem tracking and system auditing.

[0020] A further improvement of the technical solution of the present invention is that: in the data storage module, the process of storing the processed electronic accounting voucher data is:

[0021] Select a data warehouse based on the data volume, query requirements, and performance requirements, and configure the data warehouse's hardware and software environment. If the data volume is small, you can choose a lightweight relational database, such as MySQL or PostgreSQL. For medium or large data sets, choose a high-performance distributed data warehouse, such as Teradata, Greenplum, Snowflake, etc., or a data warehouse based on the Hadoop ecosystem, such as Apache Hive. If you need to support real-time queries, choose a data warehouse that supports high-concurrency access and fast response. If the query requirements are complex and need to support multi-table association, aggregation and other operations, you can choose a relational data warehouse or a distributed data warehouse that supports complex queries;

[0022] Load the cleaned and converted electronic accounting voucher data into the data warehouse, and use the parallel processing capabilities of the data warehouse to speed up data import and provide support for subsequent analysis and processing;

[0023] Partition the data in the data warehouse by time dimension, create indexes to improve query performance, and use data compression technology to reduce storage space usage and ensure data access speed;

[0024] Back up the electronic accounting voucher data in the data warehouse and store the backup data in the cloud storage server to prevent data loss caused by local disasters, and formulate detailed data recovery strategies and processes to ensure that data can be quickly restored in the event of data loss or damage;

[0025] Perform permission management for users accessing the data warehouse to ensure that users can only access data within their authorized scope, and record users' query and access behaviors for security auditing and tracking.

[0026] A further improvement of the technical solution of the present invention is that in the data integration module, the process of merging, deduplicating and associating data from different business systems is as follows:

[0027] Evaluate each business system that obtains electronic accounting voucher data, analyze its data structure, data quality, and data update frequency, extract the electronic accounting voucher data of each business system from the data warehouse, and synchronize the analysis cycle of the obtained electronic accounting voucher data;

[0028] The electronic accounting voucher data of different business systems are associated and merged through key fields to obtain the electronic voucher data set. The key fields include invoice number, invoice date, amount, payee, and payee information. The associated and merged data are verified to ensure the accuracy and consistency of the data.

[0029] Standardize the associated and merged electronic accounting voucher data to ensure data consistency and comparability, mark the key field data in the electronic voucher data set, and record the data source, structure, quality, and update frequency information.

[0030] A further improvement of the technical solution of the present invention is that in the anomaly detection module, the process of identifying abnormal data is:

[0031] According to business rules and historical data, define the standards and rules for abnormal identification, preset the types of abnormal data, namely amount abnormality, time abnormality, frequency abnormality, and set abnormal thresholds ;

[0032] Based on the associated data of the electronic voucher data set, the key features of the electronic voucher data, namely amount, time and frequency, are extracted. The abnormal data detection model is built with the machine learning algorithm to analyze potential abnormal points.

[0033] Combining the abnormal data detection model with the associated data of the electronic voucher dataset, the anomaly detection index is analyzed to identify data points within the abnormal threshold range as potential anomalies;

[0034] Further analyze the detected potential anomalies to confirm whether they are real anomalies and identify the specific types of anomalies, such as fraud, incorrect input, and system failure.

[0035] Combine the associated data and anomaly detection index of the electronic voucher data set to obtain the anomaly warning coefficient, classify the confirmed anomaly data, evaluate the severity of the anomaly, and analyze the cause of the anomaly;

[0036] Based on the abnormal warning coefficient and the evaluation results of the severity of the abnormal situation, different abnormal levels are set, namely, slight abnormal level, medium abnormal level and severe abnormal level, and the corresponding warning threshold is matched for each abnormal level.

[0037] A further improvement of the technical solution of the present invention is that the expression of the anomaly detection index is: ;

[0038] in, is the anomaly detection index, It is an electronic accounting document data set. , each data point contains multiple features, is the number of records in the dataset, is the first records, is the mean value of the data set, is the standard deviation of the data set, The value range is ,when When it is close to 1, it means the data set Records and averages in The deviation is small, the data is relatively normal, and combined with the preset abnormal threshold , to determine whether there are anomalies in the data set. If , then there are anomalies in the data set.

[0039] A further improvement of the technical solution of the present invention is that the expression of the abnormal warning coefficient is:

[0040] ;

[0041] in, is the abnormal warning coefficient, It is an electronic accounting document data set. is the number of records in the dataset, is the first records, is the mean value of the data set, is the standard deviation of the data set;

[0042] It is The weight of the record, is a tuning parameter used to control the sensitivity of the exponential function. The higher the value, the more severe the anomaly.

[0043] A further improvement of the technical solution of the present invention is that: a plurality of the abnormal levels correspond to a plurality of the warning thresholds, wherein the warning thresholds include an upper threshold and a lower threshold;

[0044] The multiple abnormal levels and the multiple warning thresholds satisfy the following relationship:

[0045] Slight abnormality level , dataset Classified as a minor abnormality;

[0046] Medium abnormality level , dataset Classified as moderately abnormal;

[0047] Severe abnormality level , dataset Classified as a severe abnormality;

[0048] in, is the abnormal warning coefficient, is the lower threshold corresponding to the medium abnormality level and the upper threshold corresponding to the slight abnormality level, It is the lower threshold corresponding to the severe abnormality level and the upper threshold corresponding to the medium abnormality level.

[0049] A further improvement of the technical solution of the present invention is that in the exception handling response module, the process of triggering the corresponding processing flow for the detected abnormal situation is:

[0050] For the detected abnormal situation, the level of the abnormal situation is determined in combination with the abnormal warning coefficient;

[0051] According to the level of abnormal situation, the corresponding processing flow is triggered. Among them, the minor abnormal level: triggers the low-priority processing flow, such as sending notification emails or system messages to relevant personnel to remind them to conduct preliminary inspections, record abnormal information, and conduct regular reviews to optimize the system; the medium abnormal level: triggers the medium-priority processing flow, including: sending more urgent notifications (such as text messages) to relevant personnel, conducting further data analysis and investigation to determine the cause of the abnormality, monitoring abnormal trends, and preparing emergency measures; the serious abnormal level: triggers the high-priority processing flow, including: taking immediate measures, such as freezing related accounts to prevent further losses, and notifying key personnel (such as system administrators, security teams) to respond immediately;

[0052] Automatically notify relevant personnel via email, SMS or system notification, mark abnormal nodes where abnormal situations occur in the data transmission path, dynamically adjust the data transmission path, bypass abnormal nodes, and ensure the continuity and stability of data transmission;

[0053] For abnormal nodes with severe abnormality levels, the abnormal data and abnormal nodes are isolated, and relevant personnel are notified to handle them first to prevent further losses and affect the normal operation of other systems;

[0054] Record the entire process of anomaly detection, level assessment, processing flow, and processing results in detail to form a complete anomaly processing log.

[0055] In the second aspect, a non-tax special account supervision method based on electronic accounting vouchers is implemented based on a non-tax special account supervision system for electronic accounting vouchers, and includes the following steps:

[0056] S1. Integrate the electronic accounting voucher data of various business systems to the supervision center, and clean and standardize the data, remove duplicates, correct errors, fill in missing values, etc., to ensure data quality and ensure that the data format, coding, classification, etc. meet regulatory requirements to facilitate subsequent analysis;

[0057] S2. Use the preset anomaly identification standards and machine learning algorithms to build an anomaly detection model, conduct in-depth analysis of electronic accounting voucher data, identify potential anomalies, and conduct further analysis and verification to confirm whether it is a real anomaly and identify the specific type of anomaly;

[0058] S3. Combining the analysis results of the anomaly detection model with the electronic accounting voucher data, classify the confirmed anomaly data and evaluate the severity of the anomaly;

[0059] S4. Classify the confirmed abnormal situations, take corresponding processing measures, dynamically adjust the data transmission path, bypass the detected abnormal nodes, and record the abnormal detection, processing process and results to form an abnormal processing log.

[0060] Due to the adoption of the above technical solution, the present invention has the following technical advances compared with the prior art:

[0061] The present invention provides a non-tax special account supervision system and method based on electronic accounting vouchers. Through automated data integration and anomaly detection mechanisms, the supervision efficiency is significantly improved, and a large amount of electronic accounting voucher data can be monitored and analyzed in real time, and abnormal situations can be quickly identified. Compared with manual review methods, it is more efficient and accurate. The automated process reduces the possibility of human errors and omissions, and also reduces the work pressure of supervisors. In addition, a built-in risk warning mechanism can immediately trigger the warning process once an abnormality or potential risk is discovered, which helps to take quick measures to prevent the risk from expanding and spreading.

[0062] The present invention provides a non-tax special account supervision system and method based on electronic accounting vouchers. By automatically integrating and processing electronic accounting voucher data, the efficiency and accuracy of supervision work can be significantly improved, data can be captured and analyzed in real time and accurately, and anomalies and potential risks can be quickly identified, making supervision work more efficient and accurate. The detection algorithm is continuously optimized through a machine learning algorithm to adapt to the ever-changing business environment and abnormal behavior patterns, thereby improving the accuracy of supervision. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required to be used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can also be obtained according to these drawings.

[0064] Figure 1 It is a module structure diagram of the present invention;

[0065] Figure 2 A flow chart for identifying abnormal data for the present invention;

[0066] Figure 3 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0068] Embodiment 1, as Figure 1 , Figure 2 As shown, the present invention provides a non-tax special account supervision system based on electronic accounting vouchers, including a supervision center, the supervision center is communicatively connected with a data integration module, a data storage module, a data integration module, an anomaly detection module, an anomaly processing response module and a report analysis module, wherein the modules are connected by electrical signals;

[0069] The data integration module is used to collect electronic accounting voucher data related to non-tax revenue from various business systems. The business systems include non-tax revenue collection system, integrated budget management system, debt management system, etc., to achieve unified access to multi-source data, provide a basis for subsequent processing, and convert and standardize the collected electronic accounting voucher data to ensure data consistency and compatibility, providing a basis for subsequent data processing and analysis. It identifies all business systems that need to be integrated, including but not limited to non-tax revenue collection system, integrated budget management system, debt management system, etc. Each business system establishes a data access protocol, accesses the electronic accounting voucher data of each business system through API interface calls, and clarifies the data type and field of electronic accounting voucher to be extracted in each business system, namely invoice number, invoice date, amount, payee, and payee information. For each business system, configure the corresponding interface docking program, such as using WebService, REST Use ETL tools to extract electronic accounting voucher data related to non-tax revenue from various business systems at predetermined time intervals through API or direct database connection, clean the acquired electronic accounting voucher data related to non-tax revenue, remove duplicates, errors and incomplete records in the data, perform preliminary data verification during data access to ensure data accuracy and completeness, establish data mapping relationships, convert data formats of different business systems into a unified format, set up a monitoring mechanism to monitor any anomalies in the data integration process, and record detailed operation logs to facilitate problem tracking and system auditing;

[0070] The data storage module uses a high-performance, high-reliability data warehouse to store the processed electronic accounting voucher data to ensure the security and integrity of the data, and provides data backup, recovery, and query functions to support rapid retrieval and access to data, and provide data support for regulatory work. The data warehouse is selected based on the data volume, query requirements, and performance requirements, and the hardware and software environment of the data warehouse is configured. If the data volume is small, a lightweight relational database such as MySQL or PostgreSQL can be selected. For medium or large data sets, a high-performance distributed data warehouse such as Teradata, Greenplum, Snowflake, etc., or a data warehouse based on the Hadoop ecosystem, such as Apache Hive: If you need to support real-time query, choose a data warehouse that supports high concurrent access and fast response. If the query requirements are complex and need to support multi-table association, aggregation and other operations, you can choose a relational data warehouse or a distributed data warehouse that supports complex queries. Load the cleaned and converted electronic accounting voucher data into the data warehouse, use the parallel processing capabilities of the data warehouse to speed up data import, and provide support for subsequent analysis and processing. Partition the data in the data warehouse by time dimension, create indexes to improve query performance, and use data compression technology to reduce storage space usage to ensure data access speed. Back up the electronic accounting voucher data in the data warehouse and store the backup data in the cloud storage server to prevent data loss caused by local disasters. Develop detailed data recovery strategies and processes to ensure that data can be quickly restored in the event of data loss or damage. Manage the permissions of users accessing the data warehouse to ensure that users can only access data within their authorized scope, and record user queries and access behaviors for security auditing and tracking.

[0071] The data integration module uses data integration technology to merge, remove duplicates and associate data from different business systems to form a unified data view, evaluate each business system that obtains electronic accounting voucher data, analyze its data structure, data quality, and data update frequency, extract the electronic accounting voucher data of each business system from the data warehouse, synchronize the analysis cycle of the acquired electronic accounting voucher data, associate and merge the electronic accounting voucher data of different business systems through key fields to obtain an electronic voucher data set. The key fields include invoice number, invoice date, amount, payee, and payer information, and verify the associated and merged data to ensure the accuracy and consistency of the data. The associated and merged electronic accounting voucher data is standardized to ensure the consistency and comparability of the data, and the key field data in the electronic voucher data set is marked to record the source, structure, quality, and update frequency information of the data;

[0072] The anomaly detection module presets the anomaly identification standards, analyzes and detects the electronic accounting voucher data, identifies the anomaly data, and further analyzes and judges the detected anomaly data to confirm whether it is a real anomaly, gives the specific anomaly type, defines the anomaly identification standards and rules based on business rules and historical data, presets the types of anomaly data, namely, amount anomaly, time anomaly, frequency anomaly, and sets the anomaly threshold Based on the associated data of the electronic voucher data set, the key features of the electronic voucher data are extracted, which are amount, time and frequency. The abnormal data detection model is constructed in combination with the machine learning algorithm to analyze the potential abnormal points. The abnormal data detection model and the associated data of the electronic voucher data set are combined to analyze and obtain the abnormal detection index. The data points within the abnormal threshold range are identified as potential abnormal points. The detected potential abnormal points are further analyzed to confirm whether they are real abnormal situations. The specific types of abnormal situations are identified, which are fraud, incorrect input, and system failure. The abnormal warning coefficient is obtained by combining the associated data of the electronic voucher data set and the abnormal detection index. The confirmed abnormal situation data is classified, and the severity of the abnormal situation is evaluated. The cause of the abnormality is analyzed. Based on the abnormal warning coefficient and the evaluation results of the severity of the abnormal situation, different abnormal levels are set, which are slight abnormal level, medium abnormal level and severe abnormal level, and the corresponding warning threshold is matched for each abnormal level;

[0073] Furthermore, the expression of the anomaly detection index is: ;

[0074] in, is the anomaly detection index, It is an electronic accounting document data set. , each data point contains multiple features, is the number of records in the dataset, is the first records, is the mean value of the data set, is the standard deviation of the data set, The value range is ,when When it is close to 1, it means that the data set is concentrated. Records and averages The deviation is small and the data is relatively normal. When it is close to 0, it means that there is a difference between the average value and the average value in the data set. Records with large deviations are abnormal. Standardize the data so that the dataset The mean is 0 and the standard deviation is 1. The influence of different dimensions and magnitudes is eliminated, and the exponential function is used. Calculate the degree of deviation of each data point from the average value, highlight outliers, and combine the preset anomaly threshold , to determine whether there are anomalies in the data set. If , then there are anomalies in the data set;

[0075] Furthermore, the expression of the abnormal warning coefficient is: ;

[0076] in, is the abnormal warning coefficient, It is an electronic accounting document data set. is the number of records in the dataset, is the first records, is the mean value of the data set, is the standard deviation of the data set, It is The weight of the record, is a tuning parameter used to control the sensitivity of the exponential function. When it is smaller, it means that the data set Records and averages in The deviation is small and the data is relatively normal. When it is larger, it means that there is a difference between the average value and the average value in the data set. Records with large deviations are abnormal. The higher the value, the higher the severity of the abnormality; further, multiple abnormality levels correspond to multiple warning thresholds, where the warning thresholds include an upper threshold and a lower threshold;

[0077] Multiple abnormality levels and multiple warning thresholds satisfy the following relationship: Minor abnormality level , dataset Classified as a minor abnormality;

[0078] Medium abnormality level , dataset Classified as moderately abnormal;

[0079] Severe abnormality level , dataset Classified as a severe abnormality;

[0080] in, is the abnormal warning coefficient, is the lower threshold corresponding to the medium abnormality level and the upper threshold corresponding to the slight abnormality level, The lower threshold corresponding to the severe abnormality level and the upper threshold corresponding to the medium abnormality level;

[0081] The exception handling response module triggers the corresponding processing flow for the detected abnormal situation, such as notifying relevant personnel, freezing relevant accounts, etc., dynamically adjusts the data transmission path, bypasses abnormal nodes, ensures the continuity and stability of data transmission, and records the exception handling process and results. For the detected abnormal situation, combined with the abnormal warning coefficient, the level of the abnormal situation is determined, and the corresponding priority is set according to the level of the abnormal situation. The priority of each abnormal level increases from minor to severe, triggering the corresponding processing flow. Among them, the minor abnormal level: triggers the low-priority processing flow, such as: sending notification emails or system messages to relevant personnel, reminding them to conduct preliminary inspections, record abnormal information, and review regularly to optimize the system; medium abnormal level: triggers the medium priority processing flow, including: sending more urgent notifications (such as text messages) to relevant personnel for further investigation. One-step data analysis and investigation to determine the cause of the anomaly, monitor the anomaly trend, and prepare emergency measures; Severe anomaly level: trigger a high-priority processing flow, including: immediate measures, such as freezing related accounts to prevent further losses, notifying key personnel (such as system administrators, security teams) to respond immediately, automatically notifying relevant personnel via email, SMS or system notification, and marking the abnormal nodes where the abnormal situation is located in the data transmission path, dynamically adjusting the data transmission path, bypassing the abnormal nodes, and ensuring the continuity and stability of data transmission. For abnormal nodes with severe anomaly levels, isolate abnormal data and abnormal nodes, and notify relevant personnel to give priority to processing to prevent further losses and affect the normal operation of other systems. Record the entire process of anomaly detection, level evaluation, processing flow, and processing results in detail to form a complete anomaly processing log;

[0082] The report analysis module is used to conduct in-depth mining and analysis of electronic accounting voucher data, generate various regulatory reports and statistical reports based on the analysis results, and provide data support.

[0083] Embodiment 2, as Figure 3 As shown, based on Example 1, the present invention also provides a non-tax special account supervision method based on electronic accounting vouchers, which is implemented based on a non-tax special account supervision system for electronic accounting vouchers and includes the following steps:

[0084] S1. Integrate the electronic accounting voucher data of various business systems to the supervision center, and clean and standardize the data, remove duplicates, correct errors, fill in missing values, etc., to ensure data quality and ensure that the data format, coding, classification, etc. meet regulatory requirements to facilitate subsequent analysis;

[0085] S2. Use the preset anomaly identification standards and machine learning algorithms to build an anomaly detection model, conduct in-depth analysis of electronic accounting voucher data, identify potential anomalies, and conduct further analysis and verification to confirm whether it is a real anomaly and identify the specific type of anomaly;

[0086] S3. Combining the analysis results of the anomaly detection model with the electronic accounting voucher data, classify the confirmed anomaly data and evaluate the severity of the anomaly;

[0087] S4. Classify the confirmed abnormal situations, take corresponding processing measures, dynamically adjust the data transmission path, bypass the detected abnormal nodes, and record the abnormal detection, processing process and results to form an abnormal processing log.

[0088] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A non-tax special account supervision system based on electronic accounting vouchers, including a supervision center, characterized by: The supervision center is communicatively connected with a data integration module, a data storage module, a data integration module, an anomaly detection module, an anomaly handling response module and a report analysis module, wherein the modules are electrically connected with each other; The data integration module is used to collect electronic accounting voucher data related to non-tax revenue from various business systems, and perform format conversion and standardization processing on the collected electronic accounting voucher data; The data storage module uses a data warehouse to store processed electronic accounting voucher data and provides data backup, recovery and query functions; The data integration module, through data integration technology, merges, removes duplication and associates data from different business systems to form a unified data view; The anomaly detection module presets anomaly identification standards, analyzes and detects electronic accounting voucher data, identifies abnormal data, and further analyzes and judges the detected abnormal data. In the anomaly detection module, the process of identifying abnormal data is: According to business rules and historical data, define the standards and rules for abnormal identification, preset the types of abnormal data, namely amount abnormality, time abnormality, frequency abnormality, and set the abnormal threshold θ; Based on the associated data of the electronic voucher data set, the key features of the electronic voucher data, namely amount, time and frequency, are extracted. The abnormal data detection model is built with the machine learning algorithm to analyze potential abnormal points. Combining the abnormal data detection model with the associated data of the electronic voucher dataset, the anomaly detection index is analyzed to identify data points within the abnormal threshold range as potential anomalies; Further analyze the detected potential anomalies to confirm whether they are real anomalies and identify the specific types of anomalies; Combine the associated data and anomaly detection index of the electronic voucher data set to obtain the anomaly warning coefficient, classify the confirmed anomaly data, evaluate the severity of the anomaly, and analyze the cause of the anomaly; Based on the abnormal warning coefficient and the evaluation results of the severity of the abnormal situation, different abnormal levels are set, namely, slight abnormal level, medium abnormal level and severe abnormal level, and the corresponding warning threshold is matched for each abnormal level. The expression of the abnormal detection index is: Where I(X) is the anomaly detection index, X is the electronic accounting voucher dataset, X = {x1, x2, ..., x n }, n is the number of records in the data set, x i is the i-th record in the data set, μ is the mean of the data set, σ is the standard deviation of the data set, and the value range of I(X) is [0, 1]; The expression of the abnormal warning coefficient is: Among them, AWC(X) is the abnormal warning coefficient, X is the electronic accounting voucher data set, n is the number of records in the data set, and x i is the i-th record in the data set, μ is the mean of the data set, σ is the standard deviation of the data set, and w i is the weight of the i-th record, k is the adjustment parameter used to control the sensitivity of the exponential function, and the higher the value of AWC(X), the higher the severity of the anomaly; The exception handling response module triggers the corresponding processing flow for the detected abnormal situation, dynamically adjusts the data transmission path, and bypasses the abnormal node; The report analysis module is used to conduct in-depth mining and analysis of electronic accounting voucher data and generate various regulatory reports and statistical reports based on the analysis results.

2. According to claim 1, a non-tax special account supervision system based on electronic accounting vouchers is characterized by: In the data integration module, the process of collecting electronic accounting voucher data related to non-tax revenue is as follows: Identify all business systems that need to be integrated, establish data access protocols for each business system, and access the electronic accounting voucher data of each business system through API interface calls; Specify the data types and fields of electronic accounting vouchers to be extracted from each business system, namely invoice number, invoice date, amount, payee, and payee information; Configure the corresponding interface docking program for each business system, and use ETL tools to extract electronic accounting voucher data related to non-tax revenue from each business system at predetermined time intervals; Clean the acquired electronic accounting voucher data related to non-tax revenue and conduct preliminary data verification during the data access process; Establish data mapping relationships, convert data formats of different business systems into a unified format, and set up a monitoring mechanism to monitor any anomalies in the data integration process and record operation logs.

3. According to claim 2, a non-tax special account supervision system based on electronic accounting vouchers is characterized by: In the data storage module, the process of storing the processed electronic accounting voucher data is as follows: Select a data warehouse based on data volume, query requirements, and performance requirements, and configure the data warehouse's hardware and software environment; Load the cleaned and converted electronic accounting voucher data into the data warehouse, and use the parallel processing capabilities of the data warehouse to speed up data import; Partition the data in the data warehouse by time dimension, create indexes to improve query performance, and use data compression technology to reduce storage space usage; Back up the electronic accounting voucher data in the data warehouse, store the backup data in the cloud storage server, and develop data recovery strategies and processes; Perform permission management for users accessing the data warehouse and record users' queries and access behaviors.

4. According to claim 3, a non-tax special account supervision system based on electronic accounting vouchers is characterized by: In the data integration module, the process of merging, deduplicating and associating data from different business systems is as follows: Evaluate each business system that obtains electronic accounting voucher data, analyze its data structure, data quality, and data update frequency, extract the electronic accounting voucher data of each business system from the data warehouse, and synchronize the analysis cycle of the obtained electronic accounting voucher data; The electronic accounting voucher data of different business systems are associated and merged through key fields to obtain the electronic voucher data set. The key fields include invoice number, invoice date, amount, payee, payee information, and the associated and merged data are verified; Standardize the associated and merged electronic accounting voucher data, mark the key field data in the electronic voucher data set, and record the source, structure, quality, and update frequency information of the data.

5. According to claim 4, a non-tax special account supervision system based on electronic accounting vouchers is characterized by: A plurality of the abnormality levels correspond to a plurality of the warning thresholds, wherein the warning thresholds include an upper threshold and a lower threshold; The multiple abnormal levels and the multiple warning thresholds satisfy the following relationship: Slight abnormality level θ1 <AWC(X)≤θ2; Medium abnormality level θ2 <AWC(X)≤θ3; Severe abnormality level AWC(X)>θ3; Among them, AWC(X) is the abnormal warning coefficient, θ2 is the lower threshold corresponding to the medium abnormal level and the upper threshold corresponding to the slight abnormal level, and θ3 is the lower threshold corresponding to the severe abnormal level and the upper threshold corresponding to the medium abnormal level.

6. A non-tax special account supervision system based on electronic accounting vouchers according to claim 5, characterized in that: In the exception handling response module, the process of triggering the corresponding processing flow for the detected abnormal situation is as follows: For the detected abnormal situation, the level of the abnormal situation is determined in combination with the abnormal warning coefficient; According to the level of abnormal situation, the corresponding priority is set. The priority of each abnormal level increases from minor to serious, triggering the corresponding processing flow. Among them, the minor abnormal level: triggers the low priority processing flow; the medium abnormal level: triggers the medium priority processing flow; the serious abnormal level: triggers the high priority processing flow; Automatically notify relevant personnel via email, SMS or system notification, mark abnormal nodes where abnormal situations occur in the data transmission path, and dynamically adjust the data transmission path to bypass abnormal nodes; For abnormal nodes with severe abnormality levels, the abnormal data and abnormal nodes are isolated, and relevant personnel are notified to handle them first; Record the entire process of anomaly detection, level assessment, processing flow, and processing results in detail to form a complete anomaly processing log.

7. A non-tax special account supervision method based on electronic accounting vouchers, implemented based on a non-tax special account supervision system based on electronic accounting vouchers as described in any one of claims 1 to 6, characterized in that: The following steps are involved: S1. Integrate the electronic accounting voucher data of various business systems to the supervision center, and clean and standardize the data; S2. Use the preset anomaly identification standards and machine learning algorithms to build an anomaly detection model, conduct in-depth analysis of electronic accounting voucher data, identify potential anomalies, and conduct further analysis and verification to confirm whether it is a real anomaly and identify the specific type of anomaly; S3. Combining the analysis results of the anomaly detection model with the electronic accounting voucher data, classify the confirmed anomaly data and evaluate the severity of the anomaly; S4. Classify the confirmed abnormal situations, take corresponding processing measures, dynamically adjust the data transmission path, bypass the detected abnormal nodes, and record the abnormal detection, processing process and results to form an abnormal processing log.

Images