Full-process end-to-end profit accounting prediction method based on big data analysis

Abstract

The invention relates to the technical field of financial data processing and business intelligent decision making, in particular to a full-process end-to-end profit accounting prediction method based on big data analysis, and the method comprises the steps: extracting full-process business operation nodes to establish a business causal map; generating a dynamic causal map; initializing profit contribution values of source nodes in the dynamic causal atlas; based on the dynamic causal atlas and in combination with a preset business conversion coefficient and a macroscopic influence weight, calculating to obtain accumulated profit contribution of the non-source node; in response to an input future macroeconomic scene, re-calculating to generate a global profit prediction result; when a deviation exists between the global profit prediction result and the reference profit, calculating a causal contribution degree index of each causal path to output a key causal path causing the deviation; the method solves the problem that the traditional method is difficult to distinguish whether the deviation is caused by the internal efficiency problem or the external environment impact.

Description

Technical Field

[0001] This invention relates to the field of financial data processing and business intelligence decision-making technology, specifically to a full-process end-to-end profit calculation and forecasting method based on big data analysis. Background Technology

[0002] In business management, profit accounting and forecasting are core aspects. Currently, companies typically rely on the analysis of financial statements to calculate profits. This method is mainly based on the accounting items of each department, collecting and summarizing historical operating results, and providing a lagging and static profit view. However, in actual operation, such traditional accounting methods have significant limitations. A company's profit generation is a dynamic process that spans multiple business segments, but existing methods struggle to accurately track and measure the contribution and loss of profit in specific, end-to-end business activities. A company's operations are profoundly affected by the external macroeconomic environment; for example, fluctuations in factors such as interest rates and consumer confidence indices directly or indirectly affect the efficiency of internal business conversion. Existing accounting models typically fail to quantify these external influences and dynamically couple them with internal business processes. Therefore, when a company's actual profit deviates from its expected target, existing technologies struggle to conduct accurate attribution analysis, failing to clearly distinguish whether the deviation is due to efficiency issues in specific internal business segments or shocks from the external macroeconomic environment. This lack of attribution ability often results in inaccurate profit forecasts and reliance on past experience rather than data-driven dynamic deduction when making forward-looking decisions such as resource allocation. This, in turn, affects the company's agility in responding to market uncertainties and the scientific nature of its decisions. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a full-process end-to-end profit calculation and prediction method based on big data analysis. Specifically, the technical solution of this invention includes: S1. Extract all business operation nodes to establish a business causal graph; calculate the macroeconomic impact weights of macroeconomic factors and causal transmission relationships in the business causal graph; combine the business causal graph with the macroeconomic impact weights to generate a dynamic causal graph. S2. Initialize the profit contribution value of the source nodes in the dynamic causal graph; based on the dynamic causal graph and combined with the preset business conversion coefficient and macroeconomic impact weight, calculate the cumulative profit contribution of the non-source nodes. S3. In response to the input future macroeconomic scenario, recalculate to generate a global profit forecast; when the global profit forecast deviates from the benchmark profit, calculate the causal contribution index of each causal path to output the key causal path that caused the deviation.

[0004] Preferably, S1 includes: S11. Extract full-process business operation nodes from heterogeneous data sources of enterprises, including enterprise resource planning systems and supply chain management systems; S12. Identify and establish causal transmission relationships between various business operation nodes through a preset business rule engine; S13. Calculate the macroeconomic impact weights using a statistical correlation analysis model.

[0005] Preferably, the macroeconomic impact weight is calculated by performing a non-linear mapping between macroeconomic factors and the Pearson correlation coefficients of business node operation indicators through a Sigmoid function to generate the macroeconomic impact weight.

[0006] Preferably, S2 includes: S21. Based on the financial accounting vouchers, initialize the profit contribution value of the source node; S22. Determine the macroeconomic impact adjustment item based on the macroeconomic impact weight; S23. The profit stream from all direct predecessor nodes will be adjusted based on the business conversion coefficient and the macroeconomic impact adjustment item. S24. Sum all the corrected profit streams to obtain the cumulative profit contribution of the current node.

[0007] Preferably, S3 includes: S31. Future macroeconomic scenarios receiving external inputs; S32. Based on future macroeconomic scenarios, trigger a recalculation of the macroeconomic impact weights and cumulative profit contributions to obtain global profit forecast results. S33. Calculate the deviation between the global profit forecast result and the benchmark profit calculated when no future macroeconomic scenario is input.

[0008] Preferably, S3 also includes: S34. Calculate the average conduction amplification factor of the causal path to be analyzed; S35. Calculate the proportion of the change in profit at the endpoint of the causal path to be analyzed in the deviation; S36. Combine the average conduction amplification coefficient with the proportion to generate the causal contribution index; S37. Sort according to the causal contribution index to output the key causal paths.

[0009] Preferably, the method further includes S4: S41. Construct a profit resilience sandbox, allowing decision-makers to perform actions including: modifying the forecast values ​​of macroeconomic factors, or adjusting the input and business conversion coefficient of the initial business nodes; S42. In response to the decision-maker's action, trigger real-time forward computation to output the global profit prediction result for that action scenario.

[0010] Preferably, S4 also includes: S43. Receive the values ​​of operating inputs before and after adjustment set by the decision-maker; S44. Based on the values ​​of operating inputs before and after the adjustment, respectively, deduce the global profit before and after the adjustment. S45. Calculate the marginal profit elasticity of operational decisions based on the values ​​of operational inputs before and after adjustment and the global profit before and after adjustment.

[0011] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention realizes refined dynamic accounting of profits in the entire process and end-to-end business activities of an enterprise. By constructing a business causal graph and calculating the profit contribution of each node, it deepens the traditional static profit collection based on accounting subjects into a traceable and measurable process calculation, accurately depicting the contribution and flow status of profits in each specific business link. 2. By constructing a macroeconomic impact weighting model, the fluctuations of the external macroeconomic environment and the conversion efficiency of the enterprise's internal business processes are dynamically and quantitatively coupled. This overcomes the shortcomings of traditional accounting models that fail to effectively integrate external impacts, and significantly improves the real-world fit and accuracy of profit accounting and forecasting models. 3. When profit forecasts deviate from benchmark values, the method can use the causal contribution index for data-driven and accurate attribution. This method can quickly locate the key business paths and macroeconomic influencing factors that cause the deviation, clearly reveal the underlying reasons for profit changes, and solve the problem that traditional methods cannot distinguish whether the deviation is caused by internal efficiency issues or external environmental shocks. 4. This invention provides a forward-looking and interactive decision support sandbox, enabling managers to conduct real-time simulations of different macroeconomic scenarios or adjust business inputs. By leveraging quantitative indicators such as marginal profit elasticity, it optimizes resource allocation, transforming profit analysis from a lagging result review into a scientific and forward-looking decision support tool, thereby enhancing the company's agility in responding to market uncertainties. Attached Figure Description

[0012] The present invention will be further explained below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0013] 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.

[0014] Example 1: Please see Figure 1 A method for end-to-end profit accounting and forecasting based on big data analysis, characterized by the following specific steps: S1. Extract all business operation nodes to establish a business causal graph; calculate the macroeconomic impact weights of macroeconomic factors and causal transmission relationships in the business causal graph; combine the business causal graph with the macroeconomic impact weights to generate a dynamic causal graph. S2. Initialize the profit contribution value of the source nodes in the dynamic causal graph; based on the dynamic causal graph and combined with the preset business conversion coefficient and macroeconomic impact weight, calculate the cumulative profit contribution of the non-source nodes. S3. In response to the input future macroeconomic scenario, recalculate to generate a global profit forecast; when the global profit forecast deviates from the benchmark profit, calculate the causal contribution index of each causal path to output the key causal path that caused the deviation.

[0015] This invention provides a full-process end-to-end profit accounting and forecasting method based on big data analysis. The method aims to build a refined profit accounting and forecasting model that can dynamically reflect the impact of internal and external factors, and realize full-link profit attribution and forward-looking forecasting from macroeconomic environment fluctuations to micro business links. The overall technical solution of this embodiment is a complete and self-consistent technical closed loop, and the specific steps include: Step S1 involves constructing a dynamic causal graph. The technical objective of this step is to build a mathematical model that depicts the interaction between internal and external business elements. This step begins by extracting business operation nodes across the entire process. These nodes are the smallest measurable units in the enterprise value chain, the atomic units of enterprise value activities, and are automatically extracted from heterogeneous data sources such as Enterprise Resource Planning (ERP), Supply Chain Management (SCM), and Customer Relationship Management (CRM). A business causal graph is established using a pre-defined business rule engine based on time series and business logic dependencies, forming causal transmission relationships between nodes. To internalize the influence of the external environment into the model, this step collects and quantifies macroeconomic factors. Furthermore, the macroeconomic impact weights of macroeconomic factors and causal transmission relationships in the business causal graph are calculated using statistical models. Finally, by combining the business causal graph with the macro-level impact weights, a dynamic causal graph is generated. This graph not only includes internal business flow relationships, but also dynamically reflects the potential impact of the external economic environment on the efficiency of each business link through weights. Step S2 involves dynamic propagation calculation of cross-departmental profit flow; the technical purpose of this step is to simulate the flow and change of profits on the constructed dynamic causal graph; this step initializes the profit contribution value of the source nodes in the dynamic causal graph based on the company's financial accounting vouchers. These source nodes are the starting point for value creation, such as a marketing investment. Based on the graph's topology, starting from the source node, the cumulative profit contribution of non-source nodes is calculated layer by layer. This calculation process is not a simple value transfer, but rather combines preset business conversion coefficients. Represents the weight of internal operational efficiency in relation to the aforementioned macroeconomic impact. It represents the adjustment of the external environment and dynamically calculates the gain or loss of profit at each stage; Step S3 involves second-level global profit forecasting and micro-attribution. The technical objective of this step is to utilize the established model for forward-looking analysis and diagnosis. This step responds to input future macroeconomic scenarios, such as a forecast of interest rates for the next quarter, triggering adjustments to the macroeconomic impact weights. and profit contribution of all nodes The recalculation was performed to generate a global profit forecast. When the global profit forecast deviates from the benchmark profit, the system traces back in reverse, calculating the causal contribution index of each causal path. Output the key causal path that leads to the bias; By deeply coupling discrete business activities within an enterprise with the external macroeconomic environment, this invention constructs a full-process, end-to-end dynamic profit simulation model. Compared with the traditional lagging and static accounting methods based on financial statements, this invention achieves dynamic, second-level prediction of future profits and can accurately attribute any changes in predicted profits to specific business paths and macroeconomic factors, providing enterprises with in-depth decision-making insights and risk response capabilities.

[0016] Example 2: S1 includes: S11. Extract full-process business operation nodes from heterogeneous data sources of enterprises, including enterprise resource planning systems and supply chain management systems; S12. Identify and establish causal transmission relationships between various business operation nodes through a preset business rule engine; S13. Calculate the macroeconomic impact weights using a statistical correlation analysis model; The calculation of macroeconomic impact weights involves using a Sigmoid function to perform a non-linear mapping between macroeconomic factors and the Pearson correlation coefficients of business node operation indicators to generate macroeconomic impact weights.

[0017] This embodiment is a concretization and optimization of step S1 in embodiment 1, aiming to ensure the accuracy, automation and sensitivity to the external environment of dynamic causal graph construction; In step S1, S11 extracts full-process business operation nodes from heterogeneous data sources of enterprises, including enterprise resource planning systems and supply chain management systems. In this embodiment, by deploying data interfaces, structured and semi-structured data are extracted from the enterprise's ERP, SCM, CRM and other databases in real time or periodically. For example, raw material batch purchase records are extracted from ERP, marketing activity placement records are extracted from CRM, and each record is defined as an independent business operation node. S12 identifies and establishes causal relationships between various business operation nodes through a pre-defined business rule engine. The pre-defined business rule engine is a software module designed to automatically construct the topology of the business causal graph. This engine has built-in rules, such as: if the business activities of node B always occur after node A in terms of timestamps, and the input data of node B, such as the order ID, comes from the output of node A, then a directed edge is established from A to B. Through this engine, the system can automatically scan all nodes and construct the initial business causal graph. To link the external environment with internal business, S13 calculates the macroeconomic impact weight using a statistical correlation analysis model. The purpose of this statistical correlation analysis model is to quantify the impact of external macroeconomic fluctuations on specific internal business processes. In this embodiment, the model is further refined as follows: the calculation of the macroeconomic impact weight involves non-linearly mapping macroeconomic factors and the Pearson correlation coefficients of business node operation indicators using a Sigmoid function to generate the macroeconomic impact weight. To clarify this design, a formula for calculating the weight of macroeconomic impacts is introduced. The purpose of this formula is to transform the abstract macroeconomic impacts into specific mathematical parameters that can directly modify the efficiency of internal business transmission. The calculation method is as follows:

[0018] in, Macroeconomic influence weights are dimensionless and calculated using this formula; their physical meaning is that of macroeconomic factors. For business processes The influence coefficient of conduction efficiency; Business Node The historical time series data of the operational metrics are numerical sequences, and their source is the company's internal business database, such as the daily exposure sequence corresponding to node advertising placements. : No. The historical time series data of each macroeconomic factor is a numerical series, and its source is a publicly available economic data provider, such as the monthly series of the consumer confidence index. The Pearson correlation coefficient calculation function outputs: The values ​​within the interval are calculated using standard statistical methods. and The Sigmoid function is a dimensionless adjustment parameter. Its value is obtained through iterative optimization by backtesting the model based on a calibrated dataset containing time series of historical business operation indicators and macroeconomic factors, aiming to minimize the historical fitting error of the model. and As a filtering parameter within the system, its value is not subjectively set, but rather extracted from the historical database. and The sequence is trained and the objective function is to minimize the root mean square error of the system output, thereby ensuring that the mapping relationship conforms to the statistical distribution law of the data itself. The formula directly relates to Pearson correlation. The range is It is not suitable for direct use as a multiplication weight; by using the Sigmoid function for nonlinear mapping, it is transformed into a... Weight values ​​within the interval This design ensures that when the correlation is high, the weight approaches 1; when the correlation is low, the weight approaches 0; and when there is no significant correlation, the weight approaches 1. The weights approach 0.5. This step uses the Sigmoid function not for simple mathematical calculations, but to solve the problem of dimensional conflicts when merging heterogeneous data sources. It maps the discrete and large fluctuation signals of external macroeconomic data into normalized control signals that are acceptable to the system through a nonlinear gating mechanism, thereby introducing external variables while ensuring system stability. To address the complex situation in the real world where a single business segment may be simultaneously affected by multiple macroeconomic factors, the model of this invention can be further extended to include situations where a single business segment... Simultaneously affected by multiple macroeconomic factors When the overall macroeconomic impact is significant, its weight is determined by the overall macroeconomic impact. By assigning weights to each individual influence To determine this, a weighted average is used. The weight It can be preset based on the experience of business experts or determined through multiple regression analysis, and This allows the model to more comprehensively reflect the combined impact of the external environment. For example, to capture the more complex nonlinear relationship between macroeconomic factors and business indicators, the Spearman rank correlation coefficient or more advanced nonlinear correlation measurement methods based on mutual information can be used to replace the Pearson correlation coefficient, thereby further improving the physical fidelity of the model. This embodiment achieves a high degree of automation and intelligence in the construction process of dynamic causal graphs. Compared with manual business process analysis, the business rule engine can construct graphs more quickly and comprehensively. Furthermore, this embodiment uses the Pearson correlation coefficient as an example to illustrate the core technical ideas and enable rapid and comprehensive graph construction in other implementations. More importantly, by introducing a macroeconomic influence weight calculation method based on the Sigmoid function, the dynamics of the graph have a solid and quantifiable mathematical foundation. The model is no longer a static internal flowchart, but a simulation system that can reflect the external economic environment in real time, greatly improving the accuracy and practical significance of subsequent predictions.

[0019] Example 3: S2 includes: S21. Based on the financial accounting vouchers, initialize the profit contribution value of the source node; S22. Determine the macroeconomic impact adjustment item based on the macroeconomic impact weight; S23. The profit stream from all direct predecessor nodes will be adjusted based on the business conversion coefficient and the macroeconomic impact adjustment item. S24. Sum all the corrected profit streams to obtain the cumulative profit contribution of the current node.

[0020] This embodiment is a detailed description of step S2 in embodiment 1, aiming to clarify the precise mathematical mechanism of the propagation calculation of profit contribution value on the dynamic causal graph; In step S2, S21 initializes the profit contribution value of the source node based on the financial accounting voucher; the source node is a node with an in-degree of zero in the dynamic causal graph, representing the initial cost input or value injection point; profit contribution value The initialization is the amount of the accounting voucher directly extracted from the enterprise's financial system and directly related to the business activities of the source node. For example, the total cost of an advertising campaign is used as a negative profit contribution or the initial profit of a direct sales order. Based on the graph's topology, starting from the source node, the cumulative profit contribution of all non-source nodes is calculated layer by layer. This process is achieved through the following steps: S22 determines the macroeconomic impact adjustment term based on the macroeconomic impact weights; the purpose of this adjustment term is to adjust the dimensionless weights calculated in the previous step. This transforms into a multiplier that can have a neutral and controllable impact on the profit stream; S23 will adjust the profit flow from all direct predecessor nodes based on the business conversion coefficient and the macroeconomic impact adjustment item; S24 sums all the corrected profit streams to obtain the cumulative profit contribution of the current node; The steps S22 to S24 described above are integrated into a unified core equation for profit flow propagation; this equation is used to compute any non-source node. Cumulative profit contribution Its derivation logic and specific form are as follows:

[0021] in, :node The cumulative profit contribution, expressed in monetary units, is calculated using this formula. : Summation symbol, representing summation of nodes All direct predecessor nodes The incoming profit streams are summarized; Direct predecessor node The cumulative profit contribution, expressed in monetary units, is calculated from the preceding steps. Business conversion coefficient, dimensionless, derived from historical data statistics of the enterprise's business systems; it represents the conversion rate from business... To business The intrinsic conversion efficiency; specifically By extracting nodes from the business system logs Flow to Node The data throughput and resource consumption records are used to calculate the physical conversion efficiency coefficient by comparing the ratio of the two. This is the macroeconomic impact adjustment item; The macro-level influence weight is dimensionless and calculated in step S1; subtracting 0.5 from it is used to centralize the influence: when That is, when macroeconomic factors are not significantly correlated, the adjustment term is 1, and has no effect; when Gain effect occurs when; A decay effect occurs over time; Global macroeconomic shock sensitivity adjustment factor : Dimensionless; its value is also based on historical data and obtained through model backtesting and calibration, used to control the overall impact of the external environment on the entire system; The core equation for profit flow propagation defined in this embodiment provides an efficient and easy-to-understand basic paradigm. In specific scenarios with extremely high accuracy requirements, the macroeconomic impact moderating term representing external influences can be upgraded to a more complex functional form, such as... To characterize internal operational efficiency Impact of external environment The potential interaction effects between them can be used to more accurately simulate complex business dynamics. By defining a clearly defined core equation for profit flow propagation, this invention elevates profit accounting from traditional departmental, outcome-based accounting subjects to a dynamic, process-based calculation at the business activity level. This equation innovatively integrates elements representing internal efficiency. and representing the external environment This means that every transmission and transformation of profits is simultaneously influenced by both internal operations and the external economy, greatly improving the precision of profit accounting and the accuracy of attribution.

[0022] Example 4: S3 includes: S31. Future macroeconomic scenarios receiving external inputs; S32. Based on future macroeconomic scenarios, trigger a recalculation of the macroeconomic impact weights and cumulative profit contributions to obtain global profit forecast results. S33. Calculate the deviation between the global profit forecast result and the benchmark profit calculated when no future macroeconomic scenario is input.

[0023] S3 also includes: S34. Calculate the average conduction amplification factor of the causal path to be analyzed; S35. Calculate the proportion of the change in profit at the endpoint of the causal path to be analyzed in the deviation; S36. Combine the average conduction amplification coefficient with the proportion to generate the causal contribution index; S37. Sort according to the causal contribution index to output the key causal paths.

[0024] This embodiment is a concretization and deepening of step S3 in embodiment 1, aiming to achieve scenario-based future profit prediction and to perform accurate and quantifiable micro-attribution when deviations occur; In step S3, the forecasting process is initiated. S31 receives external input of future macroeconomic scenarios; the future macroeconomic scenarios are a set of forecasts of key future macroeconomic factors. Forecast values, such as the GDP growth rate forecast for the next quarter published by economic analysis agencies; S32, based on future macroeconomic scenarios, triggers a recalculation of macroeconomic impact weights and cumulative profit contributions to obtain a global profit forecast; this scenario forecast will serve as input to trigger the process described in Part 1 in real time. The recalculation triggers all nodes in the second part. The recalculation ultimately yields a new global profit prediction result at the root node representing the company's total profit, which is also the endpoint of the graph. ; S33 calculates the deviation between the global profit forecast and the benchmark profit calculated without inputting future macroeconomic scenarios. Benchmark profit here It is the global profit calculated under the current or default macroeconomic scenario; when deviation When the value is not zero, the system automatically initiates attribution analysis, the specific steps of which are as follows: S34 calculates the average conduction amplification factor of the causal path to be analyzed. Causal paths to be analyzed This refers to a source node in the graph. To a certain intermediate or final business node A complete path; average conduction amplification factor The purpose is to measure the inherent ability of the path itself to amplify or reduce changes in the initial input; it is calculated as the geometric mean of the amplification coefficients of all transmission links in the path.

[0025] in, It constitutes a path One of the edges, It is a path The length is the number of sides; and These are parameter values ​​under impact scenarios; S35 calculates the proportion of the change in profit at the endpoint of the causal path to be analyzed in the deviation; the physical meaning of this proportion is the path... end Change in profits Total change in global profit The proportion in the equation is used to measure the absolute influence of the final outcome of the path. S36 combines the average conduction amplification coefficient and the proportion to generate the causal contribution index. The index is constructed using sensitivity analysis methods from systems engineering to quantify the contribution of a single path change to the overall system output change. Physically, it characterizes the energy transfer weight of a specific path to global data bias in a complex data network. Its calculation formula is as follows:

[0026] This formula combines the absolute influence of the path, i.e., the first term, and the inherent transmission properties, i.e., the second term. Furthermore, to ensure the model's universality for financial data that includes costs, i.e., negative values, a robust model without logarithmic functions was adopted. S37 sorts the key causal paths based on their causal contribution index to output the critical causal paths; the system calculates and outputs... The paths with the highest values ​​are identified as key attributions for changes in global profit, and the profit change at the endpoint of each key path is explicitly marked in the output. The direction of this path is to help decision-makers intuitively understand whether the final impact of the path is beneficial or detrimental, serving as a necessary supplementary explanation to the causal contribution index. This embodiment elevates profit analysis from the "what" level to the "why" and "how" levels; through scenario injection and real-time calculation, it achieves second-level profit prediction; and it utilizes the invented causal contribution index. It provides a data-driven attribution method that can penetrate complex business networks and accurately pinpoint which business link and which macroeconomic factor, and what kind of coupling effect led to the change in final profit, providing managers with highly in-depth decision-making support.

[0027] Example 5: The method also includes S4: S41. Construct a profit resilience sandbox, allowing decision-makers to perform actions including: modifying the forecast values ​​of macroeconomic factors, or adjusting the input and business conversion coefficient of the initial business nodes; S42. In response to the decision-maker's action, trigger real-time forward computation to output the global profit prediction result for that operation scenario; S4 also includes: S43. Receive the values ​​of operating inputs before and after adjustment set by the decision-maker; S44. Based on the values ​​of operating inputs before and after the adjustment, respectively, deduce the global profit before and after the adjustment. S45. Calculate the marginal profit elasticity of operational decisions based on the values ​​of operational inputs before and after adjustment and the global profit before and after adjustment.

[0028] This embodiment is an interactive application system provided based on the foregoing technical solution. It is a further extension of the method described in Embodiment 1, aiming to transform the computational power of the model into a tool that directly assists in management decision-making. This embodiment adds a new step S4, which includes: S41 constructs a profit resilience sandbox; the profit resilience sandbox is an interactive user interface built upon the aforementioned three core calculation modules; it allows policymakers to perform actions including: modifying macroeconomic factors. The predicted value, or adjust the investment of the initial business nodes. and business conversion coefficient For example, decision-makers can raise their future interest rate expectations from 1.5% to 2.5% on the interface, or adjust the marketing budget for product line A by one source node. Increase by 20%; S42 responds to the decision-maker's actions, triggering real-time forward computation to output the global profit prediction result for that scenario. Every parameter adjustment by the user is captured by the system and used as new input, triggering a complete and rapid forward computation from Module 1 to Module 3 in real time, and instantly refreshing the final predicted profit on the interface. ; By constructing a profit resilience sandbox, this invention encapsulates a complex back-end calculation model into an intuitive and easy-to-use what-if decision simulation tool. Managers do not need to understand the underlying complex mathematical model to instantly see the final impact of different business decisions or macroeconomic environmental changes on overall profits. This greatly reduces the technical threshold for data-driven decision-making and enhances the profit resilience of enterprises in the face of market uncertainties. To further enhance the quantitative accuracy of decision support, step S4 also includes: S43 receives the pre-adjustment value of operating inputs set by the decision-maker. Adjusted value for operating inputs For example, users set advertising spending. Ten thousand, Ten thousand; S44 derives the global profit before adjustment based on the values ​​of operating inputs before and after adjustment. With adjusted global profit The system automatically calls the core computing engine twice in the background to calculate the global profit for each of the two scenarios. S45 calculates the marginal profit elasticity of operating decisions based on the values ​​of operating inputs before and after adjustment, and the global profit before and after adjustment. The theoretical basis of this elasticity is an economic concept, and its purpose is to measure the sensitivity of the dependent variable, global profit, to changes in the independent variable, operating input. Its calculation formula is as follows:

[0029] in, The marginal profit elasticity of operational decisions is dimensionless and is calculated using this formula. The value before and after the adjustment of a certain operational investment is in currency and is set by the user in the sandbox. The total profit corresponding to the two investments is expressed in monetary units and is derived in real time by the system. The physical meaning is the percentage change in overall profit when a certain operational input changes by 1%; for example, calculating the profit of advertisement A. And the development of B This means that increasing investment in advertising A at the current stage will be more effective in improving overall profits than increasing investment in research and development B. To ensure the computational stability of the model under all inputs, the system performs a robustness check before executing the formula. Specifically, the system needs to verify the divisor; if the benchmark profit... or benchmark input The absolute value is less than a preset minimum threshold, for example If it is approximately zero, then the marginal profit elasticity is considered to be zero. If the calculation is meaningless, the system will prompt the decision-maker that the baseline value is too small and the elasticity cannot be calculated, in order to avoid program errors caused by division by zero and ensure the robustness of the model. By introducing the calculation of marginal profit elasticity, this invention provides decision-makers with a quantitative basis for optimizing resource allocation. It goes beyond simple what-if scenario simulation and can directly answer the core business question of where to allocate resources to maximize global profits among many options. This enables enterprises to shift their resource allocation decisions from relying on experience and intuition to relying on data models for precise and dynamic optimization, realizing the transformation of technological value into commercial value.

[0030] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for end-to-end profit accounting and forecasting based on big data analysis, characterized in that, The specific steps include: S1. Extract all business operation nodes to establish a business causal graph; calculate the macroeconomic impact weights of macroeconomic factors and causal transmission relationships in the business causal graph; combine the business causal graph with the macroeconomic impact weights to generate a dynamic causal graph. S2. Initialize the profit contribution value of the source nodes in the dynamic causal graph; based on the dynamic causal graph and combined with the preset business conversion coefficient and macroeconomic impact weight, calculate the cumulative profit contribution of the non-source nodes. S3. In response to the input future macroeconomic scenario, recalculate to generate a global profit forecast; when the global profit forecast deviates from the benchmark profit, calculate the causal contribution index of each causal path to output the key causal path that caused the deviation.

2. The end-to-end profit calculation and forecasting method based on big data analysis according to claim 1, characterized in that, S1 includes: S11. Extract full-process business operation nodes from heterogeneous data sources of enterprises, including enterprise resource planning systems and supply chain management systems; S12. Identify and establish causal transmission relationships between various business operation nodes through a preset business rule engine; S13. Calculate the macroeconomic impact weights using a statistical correlation analysis model.

3. The end-to-end profit calculation and forecasting method based on big data analysis according to claim 2, characterized in that, The calculation of macroeconomic impact weights involves using a Sigmoid function to perform a non-linear mapping between macroeconomic factors and the Pearson correlation coefficients of business node operation indicators to generate macroeconomic impact weights.

4. The end-to-end profit calculation and forecasting method based on big data analysis according to claim 1, characterized in that, S2 include: S21. Based on the financial accounting vouchers, initialize the profit contribution value of the source node; S22. Determine the macroeconomic impact adjustment item based on the macroeconomic impact weight; S23. The profit stream from all direct predecessor nodes will be adjusted based on the business conversion coefficient and the macroeconomic impact adjustment item. S24. Sum all the corrected profit streams to obtain the cumulative profit contribution of the current node.

5. The end-to-end profit calculation and forecasting method based on big data analysis according to claim 1, characterized in that, S3 include: S31. Future macroeconomic scenarios receiving external inputs; S32. Based on future macroeconomic scenarios, trigger a recalculation of the macroeconomic impact weights and cumulative profit contributions to obtain global profit forecast results. S33. Calculate the deviation between the global profit forecast result and the benchmark profit calculated when no future macroeconomic scenario is input.

6. The end-to-end profit calculation and forecasting method based on big data analysis according to claim 5, characterized in that, S3 also includes: S34. Calculate the average conduction amplification factor of the causal path to be analyzed; S35. Calculate the proportion of the change in profit at the endpoint of the causal path to be analyzed in the deviation; S36. Combine the average conduction amplification coefficient with the proportion to generate the causal contribution index; S37. Sort according to the causal contribution index to output the key causal paths.

7. The end-to-end profit calculation and forecasting method based on big data analysis according to claim 1, characterized in that, The method also includes S4: S41. Construct a profit resilience sandbox, allowing decision-makers to perform actions including: modifying the forecast values ​​of macroeconomic factors, or adjusting the input and business conversion coefficient of the initial business nodes; S42. In response to the decision-maker's action, trigger real-time forward computation to output the global profit prediction result for that action scenario.

8. The end-to-end profit calculation and forecasting method based on big data analysis according to claim 7, characterized in that, S4 also includes: S43. Receive the values ​​of operating inputs before and after adjustment set by the decision-maker; S44. Based on the values ​​of operating inputs before and after the adjustment, respectively, deduce the global profit before and after the adjustment. S45. Calculate the marginal profit elasticity of operational decisions based on the values ​​of operational inputs before and after adjustment and the global profit before and after adjustment.

Images