A processing system of enterprise accounting data based on a weighting algorithm

Abstract

This invention provides a system for processing enterprise accounting data based on a weighted algorithm, belonging to the field of accounting data processing technology. The system includes a data acquisition module, a matrix construction module, an anomaly identification module, and a business update module. This invention collects total revenue and total expenditure data for a complete fiscal year from an enterprise's ERP system, decomposes it according to business dimensions such as product lines and cost centers, forming revenue and expenditure characteristic data. A business dimension correlation matrix is ​​constructed, and the correlation strength between each revenue and expenditure dimension is calculated using a specific formula. Then, the coupling factor for each dimension pair is calculated, and inefficient coupling points and broken coupling points are automatically identified by comparing with industry benchmark thresholds. Finally, targeted optimization or audit instructions are generated based on the diagnostic results. The effects are monitored after execution, and new data is collected and fed back to the model. The correlation strength is updated through a weighted algorithm, thereby achieving continuous self-optimization of enterprise resource allocation.

Description

[0001] This application is a divisional application of application number 202511236867.1, filed on September 1, 2025, with the invention title "A method and system for processing enterprise accounting data" at the time of application. Technical Field

[0002] This invention relates to the field of accounting data processing technology, specifically to a system for processing enterprise accounting data based on a weighted algorithm. Background Technology

[0003] In enterprise operations management, in-depth analysis of accounting data is crucial for optimizing resource allocation and improving operational efficiency. Traditional methods mainly rely on macro-level interpretation of static reports such as income statements and balance sheets, or simple multi-dimensional drill-down analysis using business intelligence tools. These methods are widely used in scenarios such as corporate financial review, budgeting, and performance evaluation, aiming to identify operational problems from historical data. However, facing increasingly complex and detailed management needs, how to automatically and accurately identify resource mismatches and efficiency bottlenecks between specific business processes from massive amounts of accounting data, and form actionable optimization instructions, has become a core challenge for refined enterprise management.

[0004] Existing technologies typically attempt to address this issue in two ways: first, by utilizing the reporting modules built into ERP systems to summarize revenue and expenditure data according to preset dimensions, such as departments and projects, and then having analysts manually compare the differences; second, by employing independent data visualization tools to display financial indicators across different dimensions through dashboards, assisting managers in identifying anomalies. These methods are essentially retrospective and static presentations of the summarized data, and their analytical logic is based on manually set rules and thresholds.

[0005] Existing technologies have significant shortcomings. First, their analytical granularity is coarse, failing to reveal the dynamic relationship between revenue and expenditure in specific business dimensions. For example, they cannot quantify the input-output efficiency between "a certain product line" and "a specific marketing campaign." Second, they lack automated diagnostic mechanisms. The identification of anomalies relies heavily on the personal experience of managers, making it difficult to systematically discover hidden structural inefficiencies, such as imbalances in resource input-output ratios and broken connections, such as ineffective expenditures. Crucially, existing solutions are open-loop, meaning that the analysis, decision-making, and execution stages are disconnected. They cannot automatically feed the effects of optimization actions back to the analytical model, resulting in a lack of knowledge accumulation and the system's inability to continuously learn and iterate from historical management practices.

[0006] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0007] The purpose of this invention is to provide a system for processing enterprise accounting data based on a weighted algorithm, so as to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: A method for processing corporate accounting data, comprising the following specific steps: Step 1: Collect accounting data from the target company's internal ERP system for a complete fiscal year, including total expenditure and total revenue data. Decompose the accounting data according to business dimensions to obtain revenue characteristic data corresponding to at least one revenue business dimension and expenditure characteristic data corresponding to at least one expenditure business dimension. Step 2: Based on revenue feature data and expenditure feature data, construct a business dimension association matrix, where the row direction corresponds to the revenue business dimension and the column direction corresponds to the expenditure business dimension. Calculate the value of each element in the business dimension association matrix using statistical methods. This value is used to quantify the association strength between the revenue business dimension and the expenditure business dimension corresponding to its row and column. Step 3: Calculate the coupling factor of each dimension pair in the business dimension association matrix, and identify coupling anomalies based on the comparison of the coupling factor with the preset threshold. These include inefficient coupling points that indicate inefficient resource input and broken coupling points that indicate no correlation between resource input and output. Step 4: Based on the coupling anomalies, generate a resource optimization configuration instruction set for a specific business dimension, execute the instruction set, monitor its execution effect, and feed the execution effect back to the business dimension association matrix to update the association strength of the corresponding dimension pairs in the business dimension association matrix.

[0009] Furthermore, the logic for collecting accounting data from the target company's internal ERP system within a complete fiscal year is as follows: The revenue business dimension is selected from at least one of product line, sales region, and customer group; the expense business dimension is selected from at least one of cost center, project number, and supplier category; and the accounting data is the accounting data of a complete fiscal year of the previous year of the enterprise. Both the revenue characteristic data and the expense characteristic data are accompanied by traceable identifiers of the vouchers that constitute them. The construction of the business dimension association matrix is ​​further limited to: The rows of the business dimension association matrix correspond to the revenue feature data of each revenue business dimension, and the columns correspond to the expenditure feature data of each expenditure business dimension. The value of each element in the business dimension association matrix is ​​used to characterize the correlation strength between its corresponding revenue business dimension and expenditure business dimension.

[0010] Furthermore, the logic for calculating the value of each element in the business dimension association matrix using statistical methods is as follows: Elements in the business dimension association matrix , which represents the correlation strength between the i-th revenue business dimension and the j-th expenditure business dimension, and is calculated by the following formula: in, For the i-th revenue business dimension, the total revenue within a complete fiscal year. For the j-th expenditure business dimension, the total expenditure within a complete fiscal year. This is the correlation weight coefficient between the i-th revenue business dimension and the j-th expense business dimension, calculated based on historical accounting data, and it is a normalized value. It is an extremely small positive number; row index i corresponds to a certain revenue business dimension, and column index j corresponds to a certain expenditure business dimension; The , and The values ​​are all derived from the aggregated calculation results of traceable accounting voucher data for a complete fiscal year.

[0011] Furthermore, the coupling factor is calculated for each element in the business dimension association matrix, and coupling anomalies are identified and further specified as follows: Calculate the coupling factor for each dimension in the business dimension association matrix. Its mathematical expression is: in, The arithmetic mean of all elements in the i-th row. The arithmetic mean of all elements in column j is given by the given arithmetic mean. It is a very small positive number; the dimension pair is a pair of revenue business dimension i and expenditure business dimension j represented by each element in the business dimension association matrix; The calculated coupling factor Compared with the preset efficiency threshold range If a comparison is made, If so, then the dimension pair is determined to be an inefficient coupling point. If so, then the dimensional pair is determined to be a point of break coupling; Preset efficiency threshold range The determination is based on statistical analysis of publicly available benchmark efficiency data for the industry in which the company operates, and the statistical caliber used in the determination is consistent with the statistical caliber of a complete fiscal year.

[0012] Furthermore, based on coupling anomalies, the logic for generating resource optimization configuration instruction sets for specific business dimensions is as follows: For revenue-expenditure dimension pairs identified as inefficient coupling points, a first type of instruction is generated. This first type of instruction is used to reduce the cost of the j-th expenditure business dimension or increase the revenue of the i-th revenue business dimension. For revenue-expenditure dimension pairs identified as broken coupling points, a second type of instruction is generated. This second type of instruction is used to trigger a special audit of the rationality of the financial activities related to the j-th expenditure business dimension and its correlation with the i-th revenue business dimension. Each of the generated instructions is associated with its corresponding voucher list identifier and coupling factor determination basis, for execution by the business responsible department and for traceability by the audit department.

[0013] Furthermore, the logic for feeding back the execution results to the business dimension association matrix to update the association strength of the corresponding dimension pairs in the business dimension association matrix is ​​as follows: After a pre-defined full accounting period T, new accounting data generated by executing resource optimization and allocation instruction sets is collected. Based on the new accounting data, the correlation strength corresponding to the dimension pairs identified as coupling outliers is recalculated and denoted as the observed values. ; The weighted moving average algorithm is used to update the corresponding association strength in the business dimension association matrix. The update formula is as follows: in, The correlation strength before the update. For the updated association strength, To assign weighting factors to historical data, ; In the observation values Before updating, perform accounting debit and credit balance and voucher integrity verification.

[0014] Furthermore, feeding back the execution results to the business-dimensional association matrix to update the association strength includes: updating the association strength record corresponding to each update operation and the historical weight coefficient. And the corresponding voucher retrospective list is stored together. The records in the associated storage contain the following fields: association strength before update. Post-update correlation strength Observational correlation strength Update the calculation timestamp and the corresponding dimension pair identifier; the stored records are used to support version tracing and audit verification of the business dimension association matrix.

[0015] The present invention also provides a system for processing enterprise accounting data, the system being used to execute the above-described method for processing enterprise accounting data, comprising: The data acquisition module is used to collect accounting data from the target company's internal ERP system within a complete fiscal year, including total expenditure and total revenue data. It also decomposes the accounting data according to business dimensions to obtain revenue characteristic data corresponding to at least one revenue business dimension and expenditure characteristic data corresponding to at least one expenditure business dimension. The matrix construction module is used to construct a business dimension association matrix based on revenue feature data and expenditure feature data. The rows correspond to the revenue business dimension, and the columns correspond to the expenditure business dimension. The value of each element in the business dimension association matrix is ​​calculated by statistical methods. This value is used to quantify the association strength between the revenue business dimension and the expenditure business dimension corresponding to its row and column. The anomaly detection module is used to calculate the coupling factor of each dimension pair in the business dimension association matrix, and to identify coupling anomalies based on the comparison of the coupling factor with a preset threshold. These anomalies include inefficient coupling points that indicate inefficient resource input and broken coupling points that indicate no correlation between resource input and output. The business update module is used to generate resource optimization configuration instruction sets for specific business dimensions based on coupling anomalies, execute these instruction sets, monitor their execution effects, and feed the execution effects back to the business dimension association matrix to update the association strength of the corresponding dimension pairs in the business dimension association matrix.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention achieves a micro-level and dynamic mapping of the relationship between income and expenditure by refining accounting data according to business dimensions such as product lines and cost centers, and constructing a business dimension correlation matrix. This overcomes the limitation of traditional reports that can only provide macro-level static results, and enables the analysis granularity to be accurate to each specific business activity pair, thereby accurately revealing the real flow and allocation of resources between business units.

[0017] This invention utilizes a composite algorithm that integrates real-time efficiency and historical correlation to calculate correlation strength, and introduces a coupling factor to compare with industry benchmark thresholds. This enables automated and intelligent diagnosis of two types of anomalies: "inefficient coupling" and "broken coupling." This technical feature replaces subjective judgment that relies entirely on the personal experience of managers, allowing the system to continuously, objectively, and systematically identify structural efficiency losses and resource misallocation problems that are difficult for humans to detect from massive amounts of data, and generate optimization or audit instructions with clear direction.

[0018] This invention dynamically updates the business dimension correlation matrix through a weighted algorithm. This technical feature makes the entire system no longer a one-way analysis tool, but an "enterprise intelligent brain" with memory and learning capabilities. It can transform the results of each management intervention into the model's own knowledge accumulation, thereby achieving continuous self-iteration and optimization of diagnostic accuracy. This completely solves the inherent defects of traditional analysis methods, such as the disconnect between decision-making and feedback and the inability to accumulate knowledge. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 This is a line graph showing the relationship between the arithmetic mean of income and expenditure and the correlation strength of the present invention. Figure 3 This is a line graph showing the arithmetic mean of income and expenditure in this invention. Figure 4 This is a curve showing the arithmetic mean of the income of this invention versus the coupling factor. Figure 5 This is a schematic diagram of the overall system modules of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0021] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0022] Example: Please see Figures 1-4 The present invention provides a technical solution: A method for processing corporate accounting data, comprising the following specific steps: Step 1: Collect accounting data from the target company's internal ERP system for a complete fiscal year, including total expenditure and total revenue data. Decompose the accounting data according to business dimensions to obtain revenue characteristic data corresponding to at least one revenue business dimension and expenditure characteristic data corresponding to at least one expenditure business dimension. The company's consolidated accounting data is deconstructed through business dimensions and transformed into a correlation matrix with revenue as the row and expenditure as the column. Each element of this matrix is ​​a quantitative indicator used to measure the correlation strength between a certain expenditure dimension and a certain revenue dimension, supported by traceable vouchers. Technically, three key transformations were completed: First, the overall consolidated data was refined into business dimension features that can be traced back to vouchers, solving the problem of consolidated data masking business differences; second, the revenue-expenditure attribution relationship was made explicit using a matrix structure, facilitating the comparison of input-output relationships between different business dimensions; third, calculable quantitative indicators were defined on the matrix elements, facilitating subsequent anomaly identification, instruction generation, and closed-loop verification. This closed loop supports the tracking of governance measures and the adaptive updating of matrix elements. The logic for collecting accounting data from the target company's internal ERP system for a complete fiscal year is as follows: The revenue business dimension is selected from at least one of product line, sales region, and customer group; the expense business dimension is selected from at least one of cost center, project number, and supplier category; and the accounting data is the accounting data of a complete fiscal year of the previous year of the enterprise. Both the revenue characteristic data and the expense characteristic data are accompanied by traceable identifiers of the vouchers that constitute them. The data from the previous year's complete fiscal year serves as the "control group" or "baseline" for the entire analysis. The system constructs an initial business dimension correlation matrix by learning the "normal" or "historical" correlation patterns between revenue and expenditure within this baseline period. All subsequent diagnoses (coupling anomalies) are generated by comparing current or future data with this baseline pattern. A complete and fixed annual cycle, such as from January 1, 2024 to December 31, 2024, can eliminate the impact of seasonal fluctuations and different cycle lengths, ensuring that all calculations in the matrix are on the same time scale and have inherent comparability. The product line mapping rule is that the "material number" or "product code" in the voucher is associated with the "product line" category in the ERP master data; the sales region mapping rule is that the "customer number" in the voucher is associated with the "sales region" to which its registered location belongs; and the customer group mapping rule is that the "customer number" in the voucher is associated with its enterprise attributes, such as "large enterprise", "small and medium-sized enterprise", or "individual consumer". The cost center mapping rule is that the "cost center" field in the voucher is read directly; the project number mapping rule is that the "WBS element" or "project code" field in the voucher is read directly; the supplier category mapping rule is that the "supplier number" in the voucher is associated with its main product / service category, such as "raw material supplier", "software service provider", "logistics service provider"; When a voucher can be mapped to multiple dimensions, a priority must be defined. For example, for an expenditure voucher, it should be mapped to "Project Number" first, then to "Cost Center" if it does not exist, and finally to "Supplier Category". As the core platform for enterprise resource management, the accounting data of the ERP system has been reviewed and confirmed by internal business processes, and has a high degree of authenticity and completeness. Limiting the collection scope to a complete fiscal year is to ensure that the data obtained can fully cover a complete business cycle of the enterprise, thereby avoiding seasonal deviations or periodic problems caused by incomplete data, and making the subsequent analysis conclusions more representative and comparable. While total expenditures and total revenues, as aggregate indicators in financial accounting, can reflect a company's overall profit and loss situation, they cannot reveal the specific efficiency of resource allocation within the company and the relationships between business processes. Therefore, they need to be broken down to a more granular business level. The revenue breakdown is based on factors such as product line, sales region, and customer group. The product line dimension categorizes revenue according to the different products or services offered by the company, aiming to analyze the contribution and market performance of each product. The sales region dimension divides revenue by geographical area or market segment, used to assess the effectiveness of sales strategies and market potential in different regions. The customer group dimension segments revenue by customer type, size, or industry to identify core customer groups and their value contribution. In practice, companies can choose one or more dimensions in combination for analysis based on their management needs. The criteria for dividing expenditures into business dimensions include cost centers, project numbers, and supplier categories. The cost center dimension corresponds to departments or teams within the enterprise, aggregating expenditures to specific cost responsibility units to facilitate performance evaluation and cost control. The project number dimension associates expenditures with specific R&D, construction, or marketing projects for calculating project costs and return on investment. The supplier category dimension categorizes expenditures by supplier type, type of procured materials or services to analyze procurement structure and supply chain costs. Similarly, enterprises can flexibly select one or more dimensions according to their actual management needs. The decomposed revenue and expenditure characteristic data must be accompanied by traceable identifiers of the original accounting vouchers on which they were generated. This means that each summary data point can be traced back to its source accounting entry and supporting documents, such as invoices, contracts, or warehouse receipts. The decomposition process should be implemented as follows: When decomposing total revenue and total expenditure, the following steps should be followed: First, use deterministic keys for direct attribution. If the voucher contains a clear key value for the target dimension, then directly attribute the voucher amount to the corresponding dimension. Second, for vouchers that cannot be directly attributed, use allocation rules. Allocation rules can be based on the allocation ratio agreed upon in the contract, the historical proportion of the account, or the weight of the department's usage. Each allocation should record the allocation ratio and basis. Third, use text matching as a last resort. Use regular keywords and fuzzy matching to classify the voucher descriptions. Record the confidence level of the matching results for auditing purposes. Fourth, each mapping or allocation operation must be recorded in the mapping table: original voucher identifier, attribution dimension, attribution amount, attribution rule identifier, attribution timestamp, and operator or automatic task identifier. The construction of the business dimension association matrix is ​​further limited to: The rows of the business dimension association matrix correspond to the revenue feature data of each revenue business dimension, and the columns correspond to the expenditure feature data of each expenditure business dimension. The value of each element in the business dimension association matrix is ​​used to characterize the correlation strength between its corresponding revenue business dimension and expenditure business dimension.

[0023] The matrix adopts a two-dimensional table format in terms of data structure. The rows are arranged sequentially with all selected revenue business dimensions and their corresponding revenue characteristic data, while the columns are arranged sequentially with all selected expenditure business dimensions and their corresponding expenditure characteristic data. Each intersecting cell, i.e., each element, is assigned a calculated value or a state to be calculated during initialization. The ultimate purpose of this value is to quantitatively characterize the degree of correlation or resource conversion efficiency between the revenue business dimension represented by its row and the expenditure business dimension represented by its column. In short, this matrix transforms a company's financial ledger into an analytical network diagram that can map the multi-dimensional and fine-grained relationship between revenue and expenditure, providing a structured data model for in-depth diagnosis of resource allocation problems.

[0024] Step 2: Based on revenue feature data and expenditure feature data, construct a business dimension association matrix, where the row direction corresponds to the revenue business dimension and the column direction corresponds to the expenditure business dimension. Calculate the value of each element in the business dimension association matrix using statistical methods. This value is used to quantify the association strength between the revenue business dimension and the expenditure business dimension corresponding to its row and column. The logic for calculating the value of each element in the business dimension association matrix using statistical methods is as follows: The business dimension association matrix is ​​a two-dimensional structure in which rows and columns represent different business perspectives. The rows correspond to the business units or channels that generate revenue, such as a specific product line, a specific sales region, or a specific customer group. The columns correspond to the business units or projects that consume resources, such as a cost center, a specific project, or a type of supplier. Each element in the matrix The goal is to answer a key operational question with a specific numerical value: the relationship between the i-th revenue source and the j-th cost expenditure is not only a financial match, but also a quantitative manifestation of the causal or strong correlation between resource input and output efficiency. Elements in the business dimension association matrix , which represents the correlation strength between the i-th revenue business dimension and the j-th expenditure business dimension, and is calculated by the following formula: in, For the i-th revenue business dimension, the total revenue within a complete fiscal year. For the j-th expenditure business dimension, the total expenditure within a complete fiscal year. This is the correlation weight coefficient between the i-th revenue business dimension and the j-th expense business dimension, calculated based on historical accounting data, and it is a normalized value. It is an extremely small positive number; row index i corresponds to a certain revenue business dimension, and column index j corresponds to a certain expenditure business dimension; Iterate through all accounting vouchers for a complete fiscal year of the previous year. For each pair of revenue dimension i and expense dimension j, count the number of times they appear simultaneously in a single accounting voucher (or accounting transaction), and denote this as . Find all Find the maximum value in the range, and then perform max-min normalization. Through this calculation, The range of values ​​is The closer the value is to 1, the more frequent and closely related the two dimensions have been in business dealings historically. This reflects the revenue generated by a unit of expenditure. It embodies the concept of efficiency and aims to measure how much return can be obtained from the i-th revenue dimension for each unit of resources invested in expenditure j. It indicates the efficiency of resource conversion from expenditure dimension j to revenue dimension i. This ratio transforms absolute financial data (revenue and expenditure) into a relative, comparable efficiency value. The larger the ratio, the higher the efficiency of expenditure dimension j in contributing to revenue dimension i, that is, the less resources invested ( This resulted in relatively high income. This shows that the input-output ratio under this path is excellent; The larger the value, the higher the efficiency factor, showing a positive correlation, because income is the numerator, and an increase in income directly improves the efficiency value. The larger the value, the smaller the efficiency factor, showing an inverse correlation. This is because expenditure is the denominator, and an increase in expenditure dilutes the efficiency value. It is a very small positive number, and its technical significance is to prevent a certain expenditure from becoming a negative number. When the denominator is zero, the formula calculation fails, thus ensuring the mathematical rigor and stability of the formula; This section reflects the existence and stability of the relationship between the i-th revenue dimension and the j-th expenditure dimension identified based on historical data. It indicates whether the two business dimensions have had frequent and stable business transactions or resource flows throughout history. It is a weight calculated by analyzing historical accounting data, such as voucher relationships and project mapping relationships. After normalization, its value range is usually between 0 and 1, taking the natural logarithm. This is a commonly used mathematical processing method, the purpose of which is to transform linearly growing... Transforming it into a curve with a gradually slowing growth rate amplifies the smaller... The difference between values, while avoiding when When larger, it affects the final result The excessive influence of this component acts as a smoothing and compression of the range; the larger the value of this component, the stronger and more frequent the historical correlation between income dimension i and expenditure dimension j. A very high value indicates a strong correlation between these two dimensions. The value implies that historical data strongly supports the existence of a business relationship between the two; The larger, The larger the value, the stronger the positive correlation. However, due to the characteristics of the logarithmic function, The contribution of the increase to the final result is decreasing; It is the result of multiplying the two factors mentioned above, which means that the final correlation strength assessment takes into account both current efficiency and historical correlation. This comprehensive quantification, taking into account historical reliability, measures the resource conversion efficiency from expenditure j to revenue i within the current accounting period. It indicates the overall strength of a business path from resource consumption to value creation; a high [value]... The value indicates that historical data proves the existence of this path, and current data shows that this path operates very efficiently; The larger the value, the stronger the positive correlation between revenue business dimension i and expenditure business dimension j, and the more effective this correlation is not only historical but also currently. This is usually a sign of a healthy business signal. The magnitude is determined by both the efficiency factor and the historical correlation factor; a significant decrease in either factor will lower the final value. Values, for example, even if the historical correlation is strong ( Very high, but if the current efficiency is extremely low ( (very small) It won't be high either; conversely, even if a certain efficiency value is high in the current period, historical data shows that this is just a coincidence. (close to 0) It will also be adjusted to a lower level; The , and The values ​​are all derived from the aggregated calculation results of traceable accounting voucher data for a complete fiscal year, ensuring that every calculation step and every final value can be traced back to the original accounting voucher. This meets the stringent requirements of enterprise internal control and external audit for the authenticity and accuracy of data, making the entire analysis model not only a mathematical tool, but also a rigorous management decision support system.

[0025] Step 3: Calculate the coupling factor of each dimension pair in the business dimension association matrix, and identify coupling anomalies based on the comparison of the coupling factor with the preset threshold. These include inefficient coupling points that indicate inefficient resource input and broken coupling points that indicate no correlation between resource input and output. For each element in the business dimension association matrix, calculate the coupling factor and identify coupling anomalies to further refine the definition as follows: Calculate the coupling factor for each dimension in the business dimension association matrix. Its mathematical expression is: in, The arithmetic mean of all elements in the i-th row. The arithmetic mean of all elements in column j is given by the given arithmetic mean. It is a very small positive number; the dimension pair is a pair of revenue business dimension i and expenditure business dimension j represented by each element in the business dimension association matrix; This reflects the strength of the association between the i-th income dimension and the j-th expenditure dimension. The correlation strength is a relative ratio, representing the degree of deviation between the average correlation level of revenue dimension i and the average correlation level of expenditure dimension j. It measures the relative position of the correlation strength of a specific dimension pair within the entire business correlation matrix. It indicates whether the efficiency of the resource conversion path from expenditure dimension j to revenue dimension i is comparable to, significantly better than, or worse than, the average level within the enterprise. It represents the absolute correlation strength value. This is transformed into a standardized metric relative to an internal benchmark. The larger the value, the more abnormally high the contribution efficiency of expenditure dimension j to revenue dimension i among all revenue activities supported by expenditure dimension j; at the same time, the conversion efficiency of expenditure dimension j is abnormally high among all expenditures on which revenue dimension i depends. In other words, resources are disproportionately flowing to this path and generating extraordinary returns. However, in a management context, this abnormally high value often suggests excessive concentration or imbalance in resource input. It may mean that other paths that should receive resources are being ignored, or that the path itself has problems such as inaccurate cost accounting or resource crowding out. Therefore, it is judged as an inefficient coupling point. Here, inefficiency does not mean that the path itself is absolutely inefficient, but rather that from the perspective of global resource optimization, this abnormally high value is an unhealthy state and may hide the risk of efficiency loss. The larger, The larger the value, the stronger the positive correlation. This is intuitive; the higher the strength of the correlation, the higher the relative ratio tends to be. It is the average of all association strengths for income dimension i, representing the average efficiency level of income dimension i in obtaining resources from all its expenditure sources. The larger, The smaller the value, the inverse correlation, because if the overall average correlation level is high, then even if... Although the actual value is not low, its relative performance may not be outstanding. It is the average of all association strengths of expenditure dimension j, representing the average support efficacy of expenditure dimension j for all its income-contributing objects. The larger, The smaller the value, the inverse correlation occurs, for the same reason as above; Calculate the average value of the i-th row and the average of column j When calculating the coupling factor, all elements should be included, including those with a value of 0, because a value of 0 itself conveys the important information that "there is no correlation between the income and expenditure," reflecting the true structure of the business. This does not exclude marked outliers. The purpose of calculating the coupling factor is to identify anomalies relative to the current global state. Therefore, the application should include all data for calculation to maintain the objectivity of the benchmark. It is a very small positive number used to prevent the formula from failing when the average of a row or column is exactly zero, thus ensuring mathematical robustness. The specific data of some sample numbers and coupling factors are shown in Table 1.

[0026] Table 1 Data analysis reveals significant correlations between parameters across different business dimensions. The data shows a clear negative correlation between the coupling factor and the average efficiency of the revenue and expenditure dimensions. When the average efficiency of both revenue and expenditure dimensions is low, the coupling factor tends to exhibit higher values. Conversely, when the average efficiency of both dimensions increases simultaneously, the coupling factor shows a decreasing trend. This indicates that in an environment where business dimension efficiency is generally low, the relative efficiency advantage of a specific dimension pair is more readily apparent, while in an overall high-efficiency environment, the performance of each dimension pair tends to be more balanced. When analyzing the relationship between correlation strength and the average values ​​of business dimensions, the correlation strength shows a significant upward trend as the average values ​​of revenue and expenditure dimensions increase in tandem. For example, when the average values ​​of revenue and expenditure reach high levels, the corresponding correlation strength values ​​also increase to high levels; while when both average values ​​are at low levels, the correlation strength remains in a low range. This reflects the scale effect of business activities, that is, when both revenue and expenditure dimensions have a high efficiency base, the synergistic effect and value creation capability generated between them will also be enhanced accordingly. Further analysis of the relationship between coupling factor and correlation strength reveals that, despite significant fluctuations in correlation strength, coupling factor remains within a relatively stable range. This demonstrates the advantage of coupling factor as a relative evaluation indicator. By introducing an internal benchmark comparison mechanism, it effectively eliminates evaluation bias caused by differences in absolute scale, enabling business dimensions of different scales to be compared in efficiency under the same standard, thus providing a more accurate basis for enterprise resource optimization.

[0027] The calculated coupling factor Compared with the preset efficiency threshold range If a comparison is made, If so, then the dimension pair is determined to be an inefficient coupling point. If so, then the dimensional pair is determined to be a point of break coupling; like This was identified as an inefficient coupling point, meaning that the correlation strength of this dimension pair is not only abnormally high within the enterprise, but has also significantly exceeded the industry-recognized reasonable upper limit. This strongly suggests that there may be excessive waste or misconfiguration of resources invested here. The system will trigger an alarm and recommend checking whether there is excessive concentration of resources or unreasonable cost allocation. like This was identified as a point of broken coupling, meaning that the correlation strength of this dimension pair is not only relatively extremely low within the enterprise, but also significantly lower than the industry-recognized reasonable lower limit. This strongly suggests that the resource input and expected output are seriously mismatched, and the two are almost disconnected; the system will trigger an alarm, suggesting a key audit of the reasonableness and authenticity of the expenditure, as well as whether the beneficiary has been mismatched. Preset efficiency threshold range The determination was made based on statistical analysis of publicly available benchmark efficiency data of the industry in which the company operates, and the statistical caliber used in the determination was consistent with the statistical caliber of a complete fiscal year. It should be determined through statistical analysis of publicly available industry benchmark data, such as collecting data from multiple comparable companies in the industry and calculating their... The distribution of values, taking the 5th percentile as... (Lower limit), take the 95th percentile as (Upper limit), which means that 90% of companies in the industry... Value falls on Within the specified interval, anything falling outside the interval is considered abnormal. It is not arbitrarily set based on experience, but is determined after statistical analysis of publicly available benchmark efficiency data for the company's industry. This means that the efficiency calculated internally by the company will be used. The distribution of values ​​should be compared with the distribution of similar indicators of industry benchmark companies or industry averages. It is crucial to emphasize that consistency in statistical standards is essential. This means that when obtaining benchmark data from industry reports, the dimensions of revenue / expenditure classification, such as the definition of product lines and cost centers, accounting treatment methods, and calculation time windows, must be as consistent as possible with the rules used in the company's internal analysis. Only in this way can the comparison be meaningful; otherwise, it is a comparison of "apples and oranges," and the resulting thresholds will be invalid.

[0028] Step 4: Based on the coupling anomaly, generate a resource optimization configuration instruction set for a specific business dimension, execute the instruction set, monitor its execution effect, and feed the execution effect back to the business dimension association matrix to update the association strength of the corresponding dimension pairs in the business dimension association matrix. Based on coupling anomalies, the logic for generating resource optimization configuration instruction sets for specific business dimensions is as follows: For revenue-expenditure dimension pairs identified as inefficient coupling points, a first type of instruction is generated. This first type of instruction is used to reduce the cost of the j-th expenditure business dimension or increase the revenue of the i-th revenue business dimension. For revenue-expenditure dimension pairs identified as broken coupling points, a second type of instruction is generated. This second type of instruction is used to trigger a special audit of the rationality of the financial activities related to the j-th expenditure business dimension and its correlation with the i-th revenue business dimension. The first type of instruction: Inefficient coupling points represent the diminishing marginal benefits or configuration distortion of resource input. Therefore, the logic of the instruction is not simply to stop input, but to carry out precise optimization. Reducing the cost of the j-th expenditure business dimension means conducting cost structure analysis, budget reduction, or finding more economical suppliers for expenditure dimension j, such as a cost center or a project. If j is a supplier category, the aim is to reduce unnecessary resource consumption and improve input efficiency. Improving the revenue of the i-th revenue business dimension means optimizing market strategies, adjusting prices, or implementing sales incentives for revenue dimension i, such as a product line or a sales region, with the aim of generating higher revenue returns from existing resource investments. This is a cost-saving or revenue-generating optimization strategy, with the goal of adjusting an abnormally high input-output ratio to a reasonable and healthy level. The second type of instruction: The point of disconnection indicates that resource input may be completely ineffective or misallocated, or involve compliance risks. Therefore, the logic of the instruction is first and foremost audit and verification, rather than direct optimization. Triggering a special audit means initiating a targeted review process to verify the relevant financial activities of expenditure dimension j, such as the authenticity (whether it occurred), reasonableness (whether the price and quantity are fair), and relevance (whether it actually occurred to support the activities of revenue dimension i) of purchase orders, expense reimbursements, and project expenditures. This is a risk-mitigation or error-correction control strategy, the purpose of which is to identify and eliminate ineffective expenditures or correct incorrect resource mapping relationships to prevent further waste of resources. Each of the generated instructions is associated with its corresponding voucher list identifier and coupling factor determination basis, so that it can be executed by the business responsible department and traced by the audit department; This ensures the executableness and auditability of management actions. When the business department receives an instruction, it can clearly see which original vouchers (data foundation) and what calculation logic (judgment basis) the instruction is based on, so that it can understand the origin of the instruction and carry out precise operations. The audit department can trace back the entire decision-making chain to verify the rationality and compliance of management actions. The logic for feeding back the execution results to the business dimension association matrix to update the association strength of the corresponding dimension pairs in the business dimension association matrix is ​​as follows: After a pre-defined full accounting period T, new accounting data generated by executing resource optimization and allocation instruction sets is collected. Based on the new accounting data, the correlation strength corresponding to the dimension pairs identified as coupling outliers is recalculated and denoted as the observed values. ; The weighted moving average algorithm is used to update the corresponding association strength in the business dimension association matrix. The update formula is as follows: in, The correlation strength before the update. For the updated association strength, To assign weighting factors to historical data, ; The preset complete accounting period T refers to the entire time cycle from the execution of the resource optimization allocation instruction set to the collection of new data for effect evaluation and model update. This period T is a preset system parameter, and its length is usually set to a complete accounting period, such as a quarter. After executing the resource optimization allocation instruction set, such as reducing costs, adjusting strategies, and conducting audits, its effects need a certain amount of time to be reflected in the financial data. Setting T to a complete accounting period ensures that there is enough time for management actions to produce observable financial results, avoiding misjudging the policy as ineffective due to too short a time. Aligning the observation period with the company's natural accounting cycle, such as quarterly or annually, means that preliminary accounting data is always obtained at the end of the cycle, ensuring new data for feedback and learning. ) and benchmark data ( The definitions are consistent, which makes the update calculations meaningful; It reflects the latest and most reliable estimate of the correlation strength between the i-th revenue dimension and the j-th expenditure dimension after the system absorbs new observation data. It is not a simple replacement, but a smooth value that integrates historical knowledge and the latest evidence. It indicates the system's latest belief in the correlation strength of a specific business path. This belief will continue to evolve and adjust with each new data evidence, so that the business dimension correlation matrix can dynamically reflect the latest operating conditions of the enterprise. The larger the value, the more confident the system is in the strong association between the two dimensions after learning, meaning that the resource conversion path from expenditure dimension j to income dimension i is efficient and reliable. This represents the system's historical understanding prior to this update. It is the weight assigned to it. The larger, Affected by historical values The greater the influence, the more conservative the system, the less willing it is to believe new observations in a single instance, and the smoother the changes. This represents the correlation strength calculated based on new data generated after executing optimization instructions, following a complete accounting cycle T. It is the most recent evidence. It is the weight assigned to it. The larger, Subject to new evidence The greater the impact, the more sensitive the system, and the faster it can respond to changes in the business environment; It is a hyperparameter whose value determines the system's memory length and learning rate. A value close to 1, such as 0.9, indicates a system with long memory, slow updates, and the ability to effectively smooth out short-term fluctuations and noise. This makes it suitable for environments with stable business operations. A value close to 0, such as 0.1, indicates that the system has short memory, updates rapidly, and can quickly adapt to drastic changes in the business environment. However, it may also be more susceptible to random disturbances. This parameter allows enterprises to adjust it according to their own business characteristics, giving the method great flexibility. In the observation values Before being used for updates, an accounting debit-credit balance and voucher integrity check should be performed. This should be done before incorporating the voucher into the update process. Performing accounting verification acts as a "gatekeeper" to ensure the quality of the feedback learning loop. It ensures that the new evidence entering the learning process is authentic, complete, and compliant with accounting rules, preventing "garbage in, garbage out" situations and maintaining the correctness of the entire system's evolutionary direction. Feeding the execution results back to the business-dimensional association matrix to update the association strength includes: updating the association strength record for each update operation and the historical weight coefficient. And the corresponding voucher retrospective list is stored together. The records in the associated storage contain the following fields: association strength before update. Post-update correlation strength Observational correlation strength Update the calculation timestamp and the corresponding dimension pair identifier; the stored records are used to support version tracing and audit verification of the business dimension association matrix; Feedback data storage mechanisms are a crucial technical means that endow the entire data processing method with auditability and reproducibility. This goes far beyond simple data recording; it is the core infrastructure for building a trustworthy, transparent, and continuously optimizing intelligent system. Its specific technical significance and explanation are as follows: Just like an airplane's black box records every operation, this storage mechanism records the complete context of every evolution of the business dimension association matrix. This makes any change in the model output traceable, reviewable, and understandable, thereby meeting the most stringent compliance requirements of internal control and external audit. storage and This allows management to verify the effectiveness of optimization directives by comparing "expected changes" (based on the directives) with "actual changes" (based on the actual changes). This allows for the evaluation of the effectiveness of management decisions, thus completing the full management loop of "analysis-decision-execution-verification"; it also saves all input parameters, such as... Input data ( The source credentials and output results ensure that at any point in time, anyone can completely reproduce the calculation process and update results based on these stored records, thus completely eliminating the possibility of "black box" operations. Association strength before update This represents the system's understanding of the correlation strength between the i-th income dimension and the j-th expenditure dimension before this update. It serves as the baseline for calculation and the starting point for measuring the magnitude of change. Without it, the effect of this update cannot be quantified; the correlation strength cannot be observed. It represents the recalculated correlation strength value based on the new accounting data generated after the execution of optimization instructions. It is new evidence from the real world and is the original data and driving force that triggers model updates. Storing it separately can be used for ex-post auditing of the accuracy and reasonableness of the evidence itself. Post-update association strength The system represents the absorption of new evidence ( The new cognitive state formed after this is the output of the weighted moving average algorithm, and is related to... The comparison can intuitively show the direction and magnitude of the model update, which is the new state of the business dimension correlation matrix; This represents the application of historical knowledge in this algorithm update. The weights of the model are key hyperparameters that control the learning behavior of the model. This ensures the complete reproducibility of the update process; if analysis of model behavior is required, different [methods / mechanisms] can be examined. The update effect under the value; The update calculation timestamp can record the precise time when the update operation occurs, providing a temporal context. It is a key field that connects all records in chronological order, used to track the evolution of the matrix over time, and can be used for performance analysis and anomaly diagnosis. Dimension pair identifiers uniquely identify the specific dimension pair that has been updated, such as "Product Line A - Marketing Expenses". These serve as primary keys or index keys in the database, enabling efficient querying and retrieval of all historical update records for a specific dimension pair. The voucher backtracking list points to data used for calculating observations. The set of identifiers for the original accounting vouchers of that batch of new accounting data is the ultimate guarantee of data lineage and audit traceability. It directly links every data point in the model to the lowest-level business vouchers, ensuring that every link in the entire analysis chain is traceable. When it's necessary to view the state of the business dimension correlation matrix at a specific historical point in time, the system can start from the latest version and reverse-engineer each update record, utilizing the stored... and It gradually rolls back to a specified historical point in time, thereby reconstructing a matrix of historical versions. This is similar to version control systems in software engineering, such as Git. Auditors can perform the following checks by retrieving any update record: locate the original voucher through the "Voucher Retrospective List" and verify it. Whether the calculations are based on real, compliant business data; based on the stored algorithm formulas and parameters. Reproduced from , , arrive The calculation process was examined to verify its accuracy; the magnitude and direction of the update were analyzed (from...). arrive In conjunction with the "optimization instructions" executed at the time, determine whether the update aligns with business logic and management expectations.

[0029] Please see Figure 5 The present invention also provides a system for processing enterprise accounting data, the system being used to execute the above-described method for processing enterprise accounting data, comprising: The data acquisition module is used to collect accounting data from the target company's internal ERP system within a complete fiscal year, including total expenditure and total revenue data. It also decomposes the accounting data according to business dimensions to obtain revenue characteristic data corresponding to at least one revenue business dimension and expenditure characteristic data corresponding to at least one expenditure business dimension. The matrix construction module is used to construct a business dimension association matrix based on revenue feature data and expenditure feature data. The rows correspond to the revenue business dimension, and the columns correspond to the expenditure business dimension. The value of each element in the business dimension association matrix is ​​calculated by statistical methods. This value is used to quantify the association strength between the revenue business dimension and the expenditure business dimension corresponding to its row and column. The anomaly detection module is used to calculate the coupling factor of each dimension pair in the business dimension association matrix, and to identify coupling anomalies based on the comparison of the coupling factor with a preset threshold. These anomalies include inefficient coupling points that indicate inefficient resource input and broken coupling points that indicate no correlation between resource input and output. Business Update Module: Based on coupling anomalies, it generates resource optimization configuration instruction sets for specific business dimensions, executes these instruction sets, monitors their execution effects, and feeds back the execution effects to the business dimension association matrix to update the association strength of the corresponding dimension pairs in the business dimension association matrix.

[0030] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0031] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0032] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs. The above description is merely a specific implementation of this application, but the protection scope of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the protection scope of this application.

Claims

1. A system for processing enterprise accounting data based on a weighted algorithm, characterized in that, The processing system includes a data acquisition module, a matrix construction module, an anomaly detection module, and a business update module. The data acquisition module is used to collect accounting data from the target enterprise's internal ERP system within a complete fiscal year, including total expenditure and total revenue data. The accounting data is then decomposed according to business dimensions to obtain revenue characteristic data corresponding to at least one revenue business dimension and expenditure characteristic data corresponding to at least one expenditure business dimension. The matrix construction module is used to construct a business dimension association matrix based on revenue feature data and expenditure feature data, where the row direction corresponds to the revenue business dimension and the column direction corresponds to the expenditure business dimension. The value of each element in the business dimension association matrix is ​​calculated by statistical methods, and this value is used to quantify the association strength between the revenue business dimension and the expenditure business dimension corresponding to its row and column. The anomaly identification module is used to calculate the coupling factor of each dimension pair in the business dimension correlation matrix, and to identify coupling anomaly points based on the comparison of the coupling factor with a preset threshold. These anomaly points include inefficient coupling points that indicate inefficient resource input and broken coupling points that indicate no correlation between resource input and output. The business update module is used to generate a resource optimization configuration instruction set for a specific business dimension based on the coupling anomaly point, execute the instruction set, monitor its execution effect, and feed the execution effect back to the business dimension association matrix to update the association strength of the corresponding dimension pairs in the business dimension association matrix. The processing system is used to execute the following enterprise accounting data processing method, the specific steps of which include: Step 1: Collect accounting data from the target company's internal ERP system for a complete fiscal year, including total expenditure and total revenue data. Decompose the accounting data according to business dimensions to obtain revenue characteristic data corresponding to at least one revenue business dimension and expenditure characteristic data corresponding to at least one expenditure business dimension. Step 2: Based on revenue feature data and expenditure feature data, construct a business dimension association matrix, where the row direction corresponds to the revenue business dimension and the column direction corresponds to the expenditure business dimension. Calculate the value of each element in the business dimension association matrix using statistical methods. This value is used to quantify the association strength between the revenue business dimension and the expenditure business dimension corresponding to its row and column. Step 3: Calculate the coupling factor of each dimension pair in the business dimension association matrix, and identify coupling anomalies based on the comparison of the coupling factor with the preset threshold. These include inefficient coupling points that indicate inefficient resource input and broken coupling points that indicate no correlation between resource input and output. Step 4: Based on the coupling anomalies, generate a resource optimization configuration instruction set for a specific business dimension, execute the instruction set, monitor its execution effect, and feed the execution effect back to the business dimension association matrix to update the association strength of the corresponding dimension pairs in the business dimension association matrix.

2. The enterprise accounting data processing system based on a weighted algorithm according to claim 1, characterized in that: The logic for collecting accounting data from the target company's internal ERP system for a complete fiscal year is as follows: The revenue business dimension is selected from at least one of product line, sales region, and customer group; the expense business dimension is selected from at least one of cost center, project number, and supplier category; and the accounting data is the accounting data of a complete fiscal year of the previous year of the enterprise. Both the revenue characteristic data and the expense characteristic data are accompanied by traceable identifiers of the vouchers that constitute them. The construction of the business dimension association matrix is ​​further limited to: The rows of the business dimension association matrix correspond to the revenue feature data of each revenue business dimension, and the columns correspond to the expenditure feature data of each expenditure business dimension. The value of each element in the business dimension association matrix is ​​used to characterize the correlation strength between its corresponding revenue business dimension and expenditure business dimension.

3. The enterprise accounting data processing system based on a weighted algorithm according to claim 2, characterized in that: The logic for calculating the value of each element in the business dimension association matrix using statistical methods is as follows: Elements in the business dimension association matrix , which represents the correlation strength between the i-th revenue business dimension and the j-th expenditure business dimension, and is calculated by the following formula: in, For the i-th revenue business dimension, the total revenue within a complete fiscal year. For the j-th expenditure business dimension, the total expenditure within a complete fiscal year. This is the correlation weight coefficient between the i-th revenue business dimension and the j-th expense business dimension, calculated based on historical accounting data, and it is a normalized value. It is an extremely small positive number; row index i corresponds to a certain revenue business dimension, and column index j corresponds to a certain expenditure business dimension; The , and The values ​​are all derived from the aggregated calculation results of traceable accounting voucher data for a complete fiscal year.

4. The enterprise accounting data processing system based on a weighted algorithm according to claim 3, characterized in that: For each element in the business dimension association matrix, calculate the coupling factor and identify coupling anomalies to further refine the definition as follows: Calculate the coupling factor for each dimension in the business dimension association matrix. Its mathematical expression is: in, The arithmetic mean of all elements in the i-th row. The arithmetic mean of all element values ​​in column j is... It is a very small positive number; the dimension pair is a pair of revenue business dimension i and expenditure business dimension j represented by each element in the business dimension association matrix; The calculated coupling factor Compared with the preset efficiency threshold range If a comparison is made, If so, then the dimension pair is determined to be an inefficient coupling point. If so, then the dimension pair is determined to be a point of break coupling; Preset efficiency threshold range The determination is based on statistical analysis of publicly available benchmark efficiency data for the industry in which the company operates, and the statistical caliber used in the determination is consistent with the statistical caliber of a complete fiscal year.

5. The enterprise accounting data processing system based on a weighted algorithm according to claim 4, characterized in that: Based on coupling anomalies, the logic for generating resource optimization configuration instruction sets for specific business dimensions is as follows: For revenue-expenditure dimension pairs identified as inefficient coupling points, a first type of instruction is generated. This first type of instruction is used to reduce the cost of the j-th expenditure business dimension or increase the revenue of the i-th revenue business dimension. For revenue-expenditure dimension pairs identified as broken coupling points, a second type of instruction is generated. This second type of instruction is used to trigger a special audit of the rationality of the financial activities related to the j-th expenditure business dimension and its correlation with the i-th revenue business dimension. Each of the generated instructions is associated with its corresponding voucher list identifier and coupling factor determination basis, for execution by the business responsible department and for traceability by the audit department.

6. The enterprise accounting data processing system based on a weighted algorithm according to claim 5, characterized in that: The logic for feeding back the execution results to the business dimension association matrix to update the association strength of the corresponding dimension pairs in the business dimension association matrix is ​​as follows: After a pre-defined full accounting period T, new accounting data generated by executing resource optimization and allocation instruction sets is collected. Based on the new accounting data, the correlation strength corresponding to the dimension pairs identified as coupling outliers is recalculated and denoted as the observed values. ; The weighted moving average algorithm is used to update the corresponding association strength in the business dimension association matrix. The update formula is as follows: in, The correlation strength before the update. For the updated association strength, To assign weighting factors to historical data, ; In the observation values Before updating, perform accounting debit and credit balance and voucher integrity verification.

7. The enterprise accounting data processing system based on a weighted algorithm according to claim 6, characterized in that: Feeding the execution results back to the business-dimensional association matrix to update the association strength includes: updating the association strength record for each update operation and the historical weight coefficient. And the corresponding voucher retrospective list is stored together. The records in the associated storage contain the following fields: association strength before update. Post-update correlation strength Observational correlation strength Update the calculation timestamp and the corresponding dimension pair identifier; the stored records are used to support version tracing and audit verification of the business dimension association matrix.

Images