Data processing method and related device

Abstract

The embodiment of the invention provides a data processing method and a related device, and the method comprises the steps: employing a contribution value of a second data index in an interpretable logic model obtained through the operation of a preset interpretable algorithm to quantify the change reason of the numerical value of a first data index. The method comprises the steps that a plurality of data indexes are obtained, the data indexes comprise a first data index and a plurality of second data indexes, the second data indexes are influence factors influencing the change of a first numerical value, and the first numerical value is the numerical value of the first data index; obtaining an interpretable logic model according to a logic relationship between the first data index and the plurality of second data indexes, the interpretable logic model being used for obtaining a first numerical value according to the plurality of second numerical values; according to the first numerical value and the multiple second numerical values, a preset interpretable algorithm is adopted for operation to obtain contribution values of the second data indexes in the interpretable logic model, and the contribution values are used for quantifying the influence degree of the second data indexes on numerical value changes of the first data indexes.

Description

Technical Field

[0001] This application relates to the field of data analysis, and more particularly to a data processing method and related apparatus. Background Technology

[0002] Currently, the internal financial management of large enterprises is characterized by rich indicator definitions, multiple dimensions, and complex interrelationships, which brings great difficulty to the analysis of the driving forces behind the differences in indicators across periods (GAP).

[0003] Existing solutions are mainly based on analysts’ experience in different fields to decompose the GAP value that changes over time into some key business elements. For example, analysts can conclude from their business experience that the company’s revenue GAP value is attributed to factors such as shipment volume and unit price.

[0004] However, existing analytical methods can only analyze the drivers of GAP based on human experience, and cannot quantitatively analyze complex factors when the factors affecting financial indicators are complex. For example, since contribution profit is affected by multiple factors such as revenue, cost, and expenses in the same period, it is impossible to separate the impact of year-on-year changes in the R&D expense allocation rate on contribution profit. Summary of the Invention

[0005] This application provides a data processing method and related apparatus, which can use the contribution value of the second data indicator in the interpretable logic model obtained by the preset interpretable algorithm to quantify the reason for the change in the value of the first data indicator.

[0006] In view of this, firstly, this application provides a data processing method in which multiple data indicators have logical relationships, and multiple second data indicators affect the value of a first data indicator. First, multiple data indicators can be obtained, including one or more first data indicators and multiple second data indicators. The second data indicators are influencing factors affecting the change of the first value, and the first value is the value of the first data indicator. Subsequently, an interpretable logical model can be obtained based on the logical relationship between the first data indicator and the multiple second data indicators. This interpretable logical model is used to obtain the first value from the multiple second values, where the second values ​​are the values ​​of the second data indicators. Based on the first value and the multiple second values, a preset interpretable algorithm is used to calculate the contribution value of the second data indicators in the interpretable logical model. This contribution value can be used to quantify the degree of influence of the second data indicators on the value change of the first data indicator.

[0007] In this embodiment of the application, the interpretable algorithm can generally be used to explain the contribution of each feature in the machine learning model to the prediction result. Therefore, the contribution value of multiple second data indicators in the interpretable logic model can be calculated based on the interpretable algorithm, and the contribution value can be used to quantify the degree of influence of the second data indicator on the numerical change of the first data indicator, so as to realize the quantitative analysis of the reasons for the numerical change of the data indicator.

[0008] In one possible implementation, the aforementioned step of obtaining the contribution value of the second data indicator in the interpretable logic model by using a preset interpretable algorithm based on the first value and multiple second values ​​may include: obtaining a first contribution value of the second data indicator by using the Shapley Additive Model Interpretation Algorithm (SHAP) based on the first value and multiple second values, wherein the first contribution value is obtained based on the Shapley value of the second data indicator; and obtaining a second contribution value of the second data indicator by using an integral gradient algorithm based on the first value and multiple second values, wherein the second contribution value is obtained based on the integral gradient value of the second data indicator.

[0009] In this embodiment of the application, multiple contribution values ​​of multiple second data indicators can be calculated based on the SHAP algorithm and the integral gradient algorithm, so that the reasons affecting the changes in the numerical values ​​of data indicators can be quantitatively analyzed based on the contribution values ​​calculated by different algorithms.

[0010] In one possible implementation, after obtaining the contribution value of the second data indicator in the interpretable logic model by using an interpretable algorithm based on the first value and multiple second values, the method may further include: determining the attribution result of the numerical change of the first data indicator based on the multiple first contribution values ​​and multiple second contribution values, wherein the attribution result includes multiple contribution values ​​of multiple second data indicators.

[0011] In this embodiment of the application, the consistency of contribution values ​​obtained by different interpretable algorithms can be compared to determine the attribution results that affect the numerical changes of data indicators, thereby improving the consistency and reliability of the obtained attribution results.

[0012] In one possible implementation, the aforementioned attribution result for determining the numerical change of the first data indicator based on multiple first contribution values ​​and multiple second contribution values ​​may include: when the first contribution value of each second data indicator is consistent with the second contribution value, using the multiple first contribution values ​​of the multiple second data indicators or the multiple second contribution values ​​of the multiple second data indicators as the attribution result; when there is a discrepancy between the first contribution value and the second contribution value of a second data indicator, determining the attribution result based on a combination of multiple data indicators, wherein each combination of multiple data indicators includes multiple third data indicators, and the multiple third data indicators are obtained by combining multiple second data indicators.

[0013] In this embodiment of the application, when the contribution values ​​of data indicators obtained by different interpretable algorithms are consistent, the contribution value can be used to quantify and analyze the reasons for the changes in the data indicator values. When they are inconsistent, the data indicators can be processed for interaction effects and the data indicators can be recombined so that the attribution results can be determined based on the new combination of data indicators.

[0014] In one possible implementation, the aforementioned determination of attribution results based on multiple data indicator combinations may include: using the SHAP algorithm to calculate multiple third contribution values ​​for each data indicator combination, wherein the third contribution values ​​are obtained based on the shapley value of the third data indicator; using the integral gradient algorithm to calculate multiple fourth contribution values ​​for each data indicator combination, wherein the fourth contribution values ​​are obtained based on the integral gradient value of the third data indicator; determining a target data indicator combination from the multiple data indicator combinations based on the multiple third contribution values ​​and the multiple fourth contribution values, wherein the third contribution value and the fourth contribution value of each data indicator in the target data indicator combination are equal and each data indicator is independent of the others; and determining the attribution result based on the target data indicator combination.

[0015] In this embodiment of the application, when the contribution values ​​obtained by different interpretable algorithms are inconsistent, the data indicators can be recombined and a target data indicator combination with the same third and fourth contribution values ​​can be selected. This allows the attribution result to be determined based on the contribution value of each data indicator in the target data indicator combination, thereby improving the reliability of the attribution result.

[0016] In one possible implementation, the aforementioned determination of attribution results based on the combination of target data indicators may include: determining the attribution results based on the third or fourth contribution value of multiple target data indicators in the combination of target data indicators.

[0017] Secondly, this application provides a data processing apparatus, comprising:

[0018] The indicator logic expansion module is used to obtain multiple data indicators, including a first data indicator and multiple second data indicators. The multiple second data indicators are the influencing factors that affect the change of the first value, which is the value of the first data indicator.

[0019] The indicator logic encapsulation module is used to obtain an interpretable logic model based on the logical relationship between the first data indicator and multiple second data indicators. This interpretable logic model is used to obtain the first value based on multiple second values, where the second values ​​are the values ​​of the second data indicators.

[0020] The indicator interpretability analysis module is used to calculate the contribution value of the second data indicator in the interpretable logic model based on the first value and multiple second values ​​using a preset interpretable algorithm. The contribution value is used to quantify the degree of influence of the second data indicator on the numerical change of the first data indicator.

[0021] In one possible implementation, the aforementioned index interpretability analysis module is specifically used to: obtain a first contribution value of the second data index by using the Shapley Additive Model Interpretation Algorithm (SHAP) based on the first value and multiple second values, wherein the first contribution value is obtained based on the Shapley value of the second data index; and obtain a second contribution value of the second data index by using the integral gradient algorithm based on the first value and multiple second values, wherein the second contribution value is obtained based on the integral gradient value of the second data index.

[0022] In one possible implementation, after obtaining multiple contribution values ​​of multiple second data indicators in the interpretable logic model using the aforementioned interpretable algorithm, the data processing device may further include: an indicator analysis result processing and interaction effect decomposition module, used to determine the attribution result of the numerical change of the first data indicator based on the multiple first contribution values ​​and the multiple second contribution values, wherein the attribution result includes the multiple contribution values ​​of the multiple second data indicators.

[0023] In one possible implementation, the aforementioned indicator analysis result processing and interaction effect decomposition module is specifically used for: when the first contribution value of each second data indicator is consistent with the second contribution value, taking the multiple first contribution values ​​of multiple second data indicators or the multiple second contribution values ​​of multiple second data indicators as the attribution result; when there is a discrepancy between the first contribution value and the second contribution value of a second data indicator, determining the attribution result based on a combination of multiple data indicators, wherein each combination of multiple data indicators includes multiple third data indicators, and the multiple third data indicators are obtained by combining multiple second data indicators.

[0024] In one possible implementation, the aforementioned indicator analysis result processing and interaction effect decomposition module is specifically used for: using the SHAP algorithm to calculate multiple third contribution values ​​for each data indicator combination in the multiple data indicator combinations, wherein the third contribution values ​​are obtained based on the shapley value of the third data indicator; using the integral gradient algorithm to calculate multiple fourth contribution values ​​for each data indicator combination in the multiple data indicator combinations, wherein the fourth contribution values ​​are obtained based on the integral gradient value of the third data indicator; determining a target data indicator combination from the multiple data indicator combinations based on the multiple third contribution values ​​and the multiple fourth contribution values, wherein the third contribution value and the fourth contribution value of each data indicator in the target data indicator combination are equal and each data indicator is independent of each other; and determining the attribution result based on the target data indicator combination.

[0025] In one possible implementation, the aforementioned indicator analysis result processing and interaction effect decomposition module is specifically used to: determine the attribution result based on the third or fourth contribution value of multiple target data indicators in the target data indicator combination.

[0026] Thirdly, embodiments of this application provide a computing device, including a processor and a memory; the processor of at least one computing device is configured to execute instructions stored in the memory of at least one computing device to cause the computing device to perform a method as described in the first aspect or any possible implementation thereof.

[0027] Fourthly, embodiments of this application provide a computing device cluster, including at least one computing device, each computing device including a processor and a memory; the processor of the at least one computing device is used to execute instructions stored in the memory of the at least one computing device, so that the computing device cluster performs a method as described in the first aspect or any possible implementation of the first aspect.

[0028] Fifthly, embodiments of this application provide a computer-readable storage medium including computer program instructions, which, when executed by a cluster of computing devices, enable the cluster of computing devices to perform a method as described in the first aspect or any possible implementation thereof.

[0029] In a sixth aspect, embodiments of this application provide a computer program product containing instructions that, when executed by a cluster of computing devices, cause the cluster of computing devices to perform a method as described in the first aspect or any possible implementation thereof.

[0030] The technical effects of the second to sixth aspects or any of their possible implementations can be found in the first aspect or the technical effects of its related possible implementations, and will not be repeated here. Attached Figure Description

[0031] Figure 1 A schematic diagram of a system framework provided for an embodiment of this application;

[0032] Figure 2 This application provides an embodiment of a device architecture for root cause analysis of GAPs generated from multi-dimensional and multi-level data metrics.

[0033] Figure 3 A flowchart illustrating a data processing method provided in an embodiment of this application;

[0034] Figure 4 This is a schematic diagram of the structure of a data processing device provided in an embodiment of this application;

[0035] Figure 5This is a schematic diagram of the structure of a computing device provided in an embodiment of this application;

[0036] Figure 6 This is a schematic diagram of the structure of a computing device cluster provided in an embodiment of this application;

[0037] Figure 7 This is a schematic diagram of another computing device cluster provided in an embodiment of this application. Detailed Implementation

[0038] The technical solutions of the embodiments of this application will now be described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0039] In this application, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of singular or plural items. The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion, such that a process, method, system, product, or apparatus that comprises a series of units is not necessarily limited to those units, but may include other units not expressly listed or inherent to those processes, methods, products, or apparatus.

[0040] First, some concepts involved in the embodiments of this application will be introduced.

[0041] 1. SHAP Algorithm

[0042] The SHAP (shapley additive explanations) algorithm is a method that uses shapley values ​​from game theory to explain the predictions of machine learning models. Specifically, SHAP is a post-hoc explanation framework that provides a shapley value for each feature, representing its contribution to the model's predictions.

[0043] 2. Integral Gradient Algorithm

[0044] Integrated gradients are a method used in machine learning to interpret the predictions of deep neural networks. Based on the concept of gradients, they quantify the contribution of input features to the model's output. Specifically, the integrated gradient algorithm calculates the integral of the gradient path from the reference input (or baseline input) to the target input to evaluate the impact of each feature on the model's output.

[0045] In many applications, it is necessary to analyze changes in data metrics to determine the reasons for these changes. In the financial field, as more and more companies' internal financial management exhibits characteristics such as multiple dimensions of financial metrics and complex interrelationships, the analysis of the driving forces behind gaps in the changes of financial metrics over time is becoming increasingly complex.

[0046] Currently, analysis of gaps in financial indicators across time is mainly based on the experience of analysts in different fields, and quantitative analysis is not possible for complex financial indicators with multiple dimensions and levels.

[0047] Based on this, this application proposes a data analysis method that can obtain an interpretable logical model based on the logical relationship between the first data indicator and multiple second data indicators among multiple data indicators, and use a preset interpretable algorithm to calculate multiple contribution values ​​of multiple second data indicators in the interpretable logical model, and use these contribution values ​​to quantify and analyze the numerical changes of the first data indicator.

[0048] The data processing method provided in this application can be applied to root cause analysis of changes in the values ​​of data indicators, such as root cause analysis of financial indicators when gaps occur across periods.

[0049] The system architecture provided in the embodiments of this application is described below.

[0050] See Figure 1 This application provides a system architecture 100. For example... Figure 1 As shown, the system architecture may include a client 110 and a computing device cluster 120, which includes at least one computing device. Any computing device can be a terminal device, such as a desktop computer, laptop computer, or other terminal device, or it can be a server, cloud server, etc. When the computing device cluster includes multiple computing devices, the different computing devices can be of the same type or different types.

[0051] Specifically, any computing device in the computing device cluster 120 can obtain data sent by the client 110, such as data of multiple financial indicators at different times, and implement the data processing method provided in this application through the installed application or plugin.

[0052] It is worth noting that, Figure 1 The system architecture shown is merely an example and is not intended to limit its specific implementation to this example. For example, in other possible system architectures, system architecture 100 may not include the client 110, and data may be obtained directly from the data storage system by any computing device in the computing device cluster 120.

[0053] Specifically, in the embodiments of this application, the apparatus for performing root cause analysis on GAP generated from multi-dimensional and multi-level data indicators is as follows: Figure 2 As shown. The root cause analysis device 200 may include an indicator logic expansion module 201, an indicator logic optimization module 202, an indicator logic encapsulation module 203, an indicator data association module 204, an indicator interpretability analysis module 205, and an indicator analysis result processing and interaction effect decomposition module 206.

[0054] The functions of each module include:

[0055] Indicator Logic Expansion Module 201: This module is used to acquire multiple data indicators. These data indicators have complex logical relationships and can be presented as a tree structure. Based on the parent-child node logical relationships between multiple data indicators, the indicator calculation formula can be obtained. For example, if data indicators A, B, and C are acquired, the expression for the indicator calculation formula can be A = B + C based on the logical relationship between A, B, and C.

[0056] Indicator logic optimization module 202: It is used to simplify the indicator calculation formula obtained by the aforementioned indicator logic expansion module 201. For example, B = D / E, the aforementioned indicator calculation formula can be simplified to A = D / E + C.

[0057] Indicator logic encapsulation module 203: used to encapsulate the aforementioned indicator calculation formula into an interpretable logic model.

[0058] Indicator data association module 204: used to associate the data indicators obtained in the aforementioned indicator logic expansion module 201 to obtain the logical relationship between multiple data indicators.

[0059] Indicator Interpretability Analysis Module 205: Used to calculate the contribution value of data indicators in the interpretable logical model using different preset interpretable algorithms.

[0060] The indicator analysis result processing and interaction effect decomposition module 206 is used to analyze the contribution values ​​obtained from the calculation to obtain the attribution results and to process the interaction effect. That is, for the differences in different attribution results caused by non-independent features in the indicators, the interaction effect decomposition and consistency verification are performed.

[0061] The method flow provided in this application will be described below in conjunction with the aforementioned system architecture.

[0062] See Figure 3 This application provides a flowchart of a data processing method, as described below.

[0063] 301. Obtain multiple data metrics;

[0064] Data metrics are typically used to measure, evaluate, and analyze a situation or phenomenon. Examples include financial metrics, sales metrics, and operational metrics. Financial metrics may include profit margin, debt-to-equity ratio, and cash flow; sales metrics may include sales revenue and sales growth rate; and operational metrics may include production efficiency, work efficiency, and product yield rate. Specific examples are not limited here. First, multiple data metrics can be obtained. These multiple data metrics include one or more primary data metrics and multiple secondary data metrics. The secondary metrics are factors that influence changes in the values ​​of the primary data metrics.

[0065] For example, multiple data indicators can be multiple financial indicators, including product A revenue, unit price and shipment volume of sub-modules a and b, carry-over rate, adjustment rate, and module revenue share, etc. Product A revenue is the first data indicator, and the unit price and shipment volume, carry-over rate, adjustment rate, and module revenue share of sub-modules a and b are multiple second data indicators. The unit price and shipment volume, carry-over rate, adjustment rate, and module revenue share of sub-modules a and b together affect the value of product A revenue.

[0066] Furthermore, under multi-dimensional cross-management, the indicators are usually multi-dimensional composite indicators. For example, the aforementioned revenue of product A can be the revenue of product A in region A, industry B, and company C. In this case, the multi-dimensional composite indicator needs to be analyzed as a whole.

[0067] 302. Based on the logical relationship between the first data indicator and multiple second data indicators, an interpretable logical model is obtained;

[0068] In this embodiment, the acquired multiple data indicators are typically multi-dimensional and multi-level, with complex logical relationships and a tree-like structure. Therefore, based on the logical relationships between the multiple data indicators, the calculation formulas for the first data indicator and multiple second data indicators can be obtained, leading to an interpretable logical model. This interpretable logical model includes the logical relationships between the first data indicator and the multiple second data indicators. Furthermore, this interpretable logical model can obtain the corresponding value of the first data indicator based on the values ​​of the multiple second data indicators.

[0069] For example, based on the logical relationship between product A revenue and the unit price, shipment volume, carryover rate, adjustment rate, and revenue share of sub-modules a and b, the expression for calculating the indicator can be: Product A revenue = (Sub-module a unit price * Sub-module a shipment volume + Sub-module b unit price * Sub-module b shipment volume) * Carryover rate * (1 - Adjustment rate) / (1 - Revenue share of the module). Furthermore, for multi-dimensional composite indicators, they can be treated as a whole to obtain a composite indicator calculation formula, such as: [Product A revenue of region A, industry B, company C] = (Sub-module a unit price * Sub-module a shipment volume + Sub-module b unit price * Sub-module b shipment volume) * Carryover rate * (1 - Adjustment rate) / (1 - Revenue share of the module). As another example, given multiple data indicators A, B, and C, where B and C are influencing factors affecting the value of indicator A, based on the logical relationship between A, B, and C, the expression can be A = B. 2 +C. Furthermore, if indicator B is influenced by lower-level indicators D and E, and satisfies the formula B = D / E, then the aforementioned indicator calculation formula should be refined. In this case, the new indicator calculation formula can be expressed as A = (D / E). 2 +C.

[0070] After obtaining the calculation formulas for multiple data indicators, these formulas can be used as functions of an interpretable logical model to determine the interpretable logical model.

[0071] 303. The contribution value of the second data index in the interpretable logic model is obtained by using a preset interpretable algorithm.

[0072] Interpretable algorithms are commonly used in machine learning to explain the prediction results of machine learning models. Interpretable algorithms include the SHAP algorithm and the integral gradient algorithm. In this embodiment, the SHAP algorithm and the integral gradient algorithm can be used to calculate the Shapley value and integral gradient value of the second data indicator in the interpretable logic model. Based on the Shapley value and integral gradient value of the second data indicator, multiple contribution values ​​of multiple second data indicators can be determined. These contribution values ​​are used to quantify the degree of influence of the second data indicator on the numerical change of the first data indicator.

[0073] Specifically, based on the value of the first data indicator (first value) and the values ​​of multiple second data indicators (second values), the SHAP algorithm and the integral gradient algorithm can be used to calculate the Shapley value and integral gradient value of each second data indicator in the model, respectively. Furthermore, the contribution value of each second data indicator can be calculated based on the Shapley value or integral gradient value of each second data indicator. This contribution value can represent the contribution of each second data indicator to the difference in the value of the first data indicator (the output value of the model). The sum of the contribution values ​​of each second data indicator is equal to the difference between the actual value and the predicted value of the model output. For example, the average predicted price of an apartment is 310,000 yuan, while the actual predicted price of a certain apartment (50 square meters, located on the 2nd floor, near a park, and no cats allowed) is 300,000 yuan, a difference of 10,000 yuan from the average predicted price. Known features affecting the apartment price include surrounding facilities, whether cats are allowed, size, and floor level. Based on the SHAP algorithm or integral gradient algorithm, the contribution value of each feature can be calculated. For example, the contribution value of "near a park" is 30,000 yuan, "50 square meters" is 10,000 yuan, "located on the 2nd floor" is 0 yuan, and "no cats allowed" is -50,000 yuan. The total contribution value is -10,000 yuan, which equals the difference between the actual predicted price and the average predicted price of the apartment.

[0074] Optionally, the SHAP algorithm can be used to calculate multiple first contribution values ​​of multiple second indicator data, which are obtained based on the shapley value of the second data indicator. Alternatively, the integral gradient algorithm can be used to calculate multiple second contribution values ​​of multiple second indicator data, which are obtained based on the integral gradient value of the second data indicator.

[0075] The operational expression for the Shapley value of each second data metric in the interpretable logical model can be:

[0076]

[0077] Where, φ j(val) represents the Shapley value of the j-th second indicator data, and S represents the value of x removed from the multiple second indicator data. j A subset of , x is a vector of second data indicators, M is the number of second data indicators, |S| is the number of second data indicators in the subset, val(S∪x) j ) are subsets S and x j The value (or benefit) created by cooperation, val(S) is the value (or benefit) created by cooperation of a subset S.

[0078] The expression for the output value of the interpretable logic model obtained based on a certain sample data (which includes the values ​​of a first data indicator and multiple second data indicators) can be:

[0079]

[0080] in, It can be regarded as a fixed item. The Shapley value of indicator x (any second data indicator). The Shapley value of the index y. denoted as the Shapley value of the index z.

[0081] Therefore, the output value corresponding to a certain sample data can be the sum of the Shapley values ​​of multiple data indicators. The difference between the values ​​of the first data indicator in different sample data can be the sum of the first contribution values ​​of multiple second data indicators. The first contribution value of the second data indicator can be the difference between the Shapley value obtained by the second data indicator based on the first sample data and the Shapley value obtained based on the second sample data. Its expression can be:

[0082]

[0083] in, The first contribution value of indicator x, The first contribution value of indicator y, This is the first contribution value of indicator z.

[0084] Furthermore, the expression for the integral gradient value of each second data index in the interpretable logical model can be:

[0085]

[0086] At this point, the expression for the output value of the interpretable logic model obtained based on a certain sample data can be:

[0087]

[0088] At this point, the difference between the values ​​of the first data indicator in different sample data can be the sum of the second contribution values ​​of multiple second data indicators. The second contribution value of each second data indicator can be the difference between the integral gradient value obtained from the first sample data and the integral gradient value obtained from the second sample data, and its expression can be:

[0089]

[0090] in, The second contribution value of indicator x, The second contribution value of indicator y, This is the second contribution value of the indicator z.

[0091] Optionally, the attribution result of the numerical change of the first data indicator can be determined based on multiple first contribution values ​​and multiple second contribution values, and the attribution result includes multiple contribution values ​​of multiple second data indicators.

[0092] In this embodiment, the SHAP algorithm and the integral gradient algorithm can be used to calculate multiple contribution values ​​of multiple second data indicators, so that the consistency of contribution values ​​obtained by different algorithms can be compared subsequently, the attribution results affecting the numerical changes of data indicators can be determined, thereby improving the consistency and reliability of the attribution results.

[0093] Specifically, after obtaining multiple first contribution values ​​and multiple second contribution values, it can be determined whether the first contribution value and the second contribution value corresponding to each second data indicator are consistent. If the first contribution value and the second contribution value of each second data indicator are consistent, then the multiple first contribution values ​​or the multiple second contribution values ​​of the multiple second data indicators can be used as the attribution result. If there are inconsistencies between the first contribution values ​​and the second contribution values ​​of the second data indicators, then the multiple second data indicators can be recombine to obtain multiple data indicator combinations. Each data indicator combination includes multiple third data indicators. Subsequently, the attribution result can be determined based on the multiple data indicator combinations.

[0094] Alternatively, when the first contribution value and the second contribution value of a second data indicator are inconsistent, after expert analysis and judgment, multiple first contribution values ​​obtained based on the SHAP algorithm can be directly selected as the contribution values ​​of multiple second data indicators to the numerical change of the first data indicator.

[0095] In this embodiment, since the integral gradient algorithm obtains the corresponding integral gradient value by finding an integral path, when the data index is not suitable for solving by a straight integral path, using the integral gradient algorithm will result in an unfair distribution of the contribution to the data index. Therefore, when the first contribution value and the second contribution value of the second data index are inconsistent, the first contribution value obtained by the SHAP algorithm can be selected as the contribution value of the data index.

[0096] When the first and second contribution values ​​of a second data indicator are inconsistent, the SHAP algorithm and the integral gradient algorithm can be used to calculate multiple Shapley values ​​and multiple integral gradient values ​​for multiple data indicator combinations. Multiple third contribution values ​​are then obtained from the multiple Shapley values, and multiple fourth contribution values ​​are obtained from the multiple integral gradient values. Further, by determining whether the multiple third and fourth contribution values ​​of the multiple data indicators are consistent, a target data indicator combination is determined from the multiple data indicator combinations, and the attribution result is determined based on this target data indicator combination. If there is no target data indicator combination among the multiple data indicator combinations where the third and fourth contribution values ​​are consistent, then the multiple first contribution values ​​of the multiple second data indicators obtained by the SHAP algorithm can be selected as the contribution values ​​of the multiple second data indicators to the numerical change of the first data indicator.

[0097] Specifically, the SHAP algorithm can be used to calculate multiple third contribution values ​​for each data indicator combination. These third contribution values ​​are obtained based on the shapley value of the third data indicator; that is, the third contribution value can be the difference between the shapley value of the third data indicator based on the first data sample and the shapley value based on the second data sample. The integral gradient algorithm can then be used to calculate multiple fourth contribution values ​​for each data indicator combination. These fourth contribution values ​​are obtained based on the integral gradient value of the third data indicator; that is, the fourth contribution value can be the difference between the integral gradient value of the third data indicator based on the first data sample and the integral gradient value based on the second data sample. By comparing the third and fourth contribution values ​​of the third data indicator in each data indicator combination, a target data indicator combination is determined from the multiple data indicator combinations. In this target data indicator combination, the shapley value and integral gradient value of each data indicator are equal, and each data indicator is independent of the others.

[0098] After determining the target data indicator combination, the attribution results affecting the numerical change of the first data indicator can be obtained based on the third and fourth contribution values ​​of multiple target data indicators in the target data combination. Specifically, the contribution values ​​of multiple second data indicators to the first data indicator can be determined based on the third or fourth contribution values ​​of multiple target data indicators, thereby obtaining the attribution results affecting the numerical change of the first data indicator.

[0099] For example, multiple data indicators X, Y, and Z are obtained, where Z is the first data indicator, and X and Y are the second data indicators. Based on the logical relationship between the data indicators X, Y, and Z, the function expression of the interpretable logical model is obtained as Z = f(X,Y) = X + 2XY. 2 +3Y+1, at this point, X is 3 and Y is 4 in sample 0, and X is 6 and Y is 10 in sample 1. First, the first and second contribution values ​​of X and Y are calculated using the SHAP algorithm and the integral gradient algorithm, respectively. When the first and second contribution values ​​of X and Y are inconsistent, X and Y can be recombine to obtain multiple data indicator combinations. Data indicator combination 1 includes data indicators X, Y, and XY; data indicator combination 2 includes data indicators X, Y, and YY. 2 Data indicator combination 3 includes data indicators X, Y, and XY. 2 Subsequently, based on sample 0 and sample 1, the Shapley value and integral gradient value of each of the aforementioned three data indicator combinations can be calculated. Then, based on the Shapley value and integral gradient value obtained for each data indicator based on different sample data, the third contribution value and fourth contribution value corresponding to each data indicator can be calculated. The third contribution value represents the contribution of X to the difference in f(X,Y) based on different sample data, and the fourth contribution value represents the contribution of Y to the difference in f(X,Y) based on different sample data. The specific calculation results are shown below.

[0100] The third and fourth contribution values ​​of each third data indicator in data indicator combination 1 are: The third and fourth contribution values ​​of each third data indicator in data indicator combination 2 are: The third and fourth contribution values ​​of each third data indicator in data indicator combination 3 are:

[0101] The calculation results show that the third and fourth contribution values ​​of each data indicator in the three data indicator combinations are consistent. Furthermore, due to the feature independence requirement of the SHAP algorithm, the data indicators in data indicator combination 2 are independent of each other. Therefore, data indicator combination 2 can be used as the target data indicator combination. Based on the target data indicators X, Y, and Y' in data indicator combination 2... 2 The contribution values ​​of the X and Y indicators are shown in Table 1.

[0102] Table 1

[0103]

[0104] As shown in Table 1, the difference between the output values ​​(the values ​​of the first data indicator, i.e., the first values) corresponding to Sample 0 and Sample 1 is Δ = f(X1,Y1) - f(X0,Y0) = 1237 - 112 = 1125. Based on the contribution value of the third data indicator in data indicator combination 2, the contribution value of indicator X to the difference between the output values ​​of Sample 0 and Sample 1 is 351, and the contribution value of indicator Y to the difference between the output values ​​of Sample 0 and Sample 1 is 774.

[0105] In this embodiment, an interpretable algorithm can be used to calculate the contribution values ​​of multiple second data indicators in the interpretable logic model. These contribution values ​​are then used to quantify the influence of the multiple second data indicators on the numerical changes of the first data indicator, thereby achieving a quantitative analysis of the causes of data indicator numerical changes. Furthermore, two different interpretable algorithms can be used to calculate the contribution values ​​of the data indicators, and the attribution result can be determined based on the consistency of the contribution values, thus improving the consistency and reliability of the attribution result.

[0106] The method flow provided in this application has been described above. The apparatus provided in this application will now be described based on the aforementioned method flow.

[0107] See Figure 4 The present application provides a schematic diagram of the structure of a data processing device, as shown below.

[0108] The indicator logic expansion module 201 is used to obtain multiple data indicators, including a first data indicator and multiple second data indicators. The multiple second data indicators are the influencing factors that affect the change of the first value, and the first value is the value of the first data indicator.

[0109] The indicator logic encapsulation module 203 is used to obtain an interpretable logic model based on the logical relationship between the first data indicator and multiple second data indicators. The interpretable logic model is used to obtain the first value based on multiple second values, where the second value is the value of the second data indicator.

[0110] The indicator interpretability analysis module 205 is used to calculate the contribution value of the second data indicator in the interpretable logic model based on the first value and multiple second values ​​using a preset interpretable algorithm. The contribution value is used to quantify the degree of influence of the second data indicator on the numerical change of the first data indicator.

[0111] In one possible implementation, the aforementioned index interpretability analysis module 205 is specifically used to: calculate a first contribution value of the second data index using the Shapley Additive Model Interpretation Algorithm (SHAP) based on the first value and multiple second values, wherein the first contribution value is obtained based on the Shapley value of the second data index; and calculate a second contribution value of the second data index using the integral gradient algorithm based on the first value and multiple second values, wherein the second contribution value is obtained based on the integral gradient value of the second data index.

[0112] In one possible implementation, after obtaining multiple contribution values ​​of multiple second data indicators in the interpretable logic model using the aforementioned interpretable algorithm, the data processing device may further include: an indicator analysis result processing and interaction effect decomposition module 206, used to determine the attribution result of the numerical change of the first data indicator based on the multiple first contribution values ​​and the multiple second contribution values, wherein the attribution result includes the multiple contribution values ​​of the multiple second data indicators.

[0113] In one possible implementation, the aforementioned indicator analysis result processing and interaction effect decomposition module 206 is specifically used for: when the first contribution value of each second data indicator is consistent with the second contribution value, taking the multiple first contribution values ​​of multiple second data indicators or the multiple second contribution values ​​of multiple second data indicators as the attribution result; when there is a discrepancy between the first contribution value and the second contribution value of a second data indicator, determining the attribution result based on the combination of multiple data indicators, wherein each data indicator combination in the multiple data indicator combination includes multiple third data indicators, and the multiple third data indicators are obtained by combining multiple second data indicators.

[0114] In one possible implementation, the aforementioned index analysis result processing and interaction effect decomposition module 206 is specifically used for: using the SHAP algorithm to calculate multiple third contribution values ​​for each data index combination in the multiple data index combinations, wherein the third contribution values ​​are obtained based on the shapley value of the third data index; using the integral gradient algorithm to calculate multiple fourth contribution values ​​for each data index combination in the multiple data index combinations, wherein the fourth contribution values ​​are obtained based on the integral gradient value of the third data index; determining a target data index combination from the multiple data index combinations based on the multiple third contribution values ​​and the multiple fourth contribution values, wherein the third contribution value and the fourth contribution value of each data index in the target data index combination are equal and each data index is independent of each other; and determining the attribution result based on the target data index combination.

[0115] In one possible implementation, the aforementioned indicator analysis result processing and interaction effect decomposition module 206 is specifically used to: determine the attribution result based on the third or fourth contribution value of multiple target data indicators in the target data indicator combination.

[0116] The indicator logic expansion module, indicator logic encapsulation module, indicator interpretability analysis module, indicator analysis result processing module, and interaction effect decomposition module can all be implemented in software or hardware. For example, the implementation of the indicator logic expansion module will be described below. Similarly, the implementation methods of the indicator logic encapsulation module, indicator interpretability analysis module, and indicator analysis result processing and interaction effect decomposition module can refer to the implementation method of the indicator logic expansion module.

[0117] As an example of a software functional unit, a module, specifically an indicator logic expansion module, may include code running on a computing instance. A computing instance may include at least one of a physical host (computing device), a virtual machine, or a container. Furthermore, the aforementioned computing instance may be one or more. For example, module A may include code running on multiple hosts / virtual machines / containers. It should be noted that the multiple hosts / virtual machines / containers used to run the code can be distributed within the same region or in different regions. Further, the multiple hosts / virtual machines / containers used to run the code can be distributed within the same availability zone (AZ) or in different AZs, each AZ comprising one or more geographically proximate data centers. Typically, a region may include multiple AZs.

[0118] Similarly, multiple hosts / virtual machines / containers used to run this code can be distributed within the same Virtual Private Cloud (VPC) or across multiple VPCs. Typically, a VPC is set up within a region. Communication between two VPCs within the same region, as well as between VPCs in different regions, requires a communication gateway to be set up within each VPC to enable interconnection between VPCs.

[0119] As an example of a hardware functional unit, the indicator logic expansion module may include at least one computing device, such as a server. Alternatively, the indicator logic expansion module may be implemented using a central processing unit (CPU), an application-specific integrated circuit (ASIC), or a programmable logic device (PLD). The aforementioned PLD may be implemented using a complex programmable logical device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), a data processing unit (DPU), a neural network processing unit (NPU), a system-on-chip (SoC), an offload card, an accelerator card, or any combination thereof.

[0120] The multiple computing devices included in the indicator logic expansion module can be distributed within the same region or in different regions. Similarly, they can be distributed within the same Availability Zone (AZ) or in different AZs. Likewise, they can be distributed within the same Virtual Private Cloud (VPC) or multiple VPCs. These multiple computing devices can be any combination of computing devices such as servers, ASICs, PLDs, CPLDs, FPGAs, GALs, DPUs, NPUs, SoCs, offloading cards, and accelerator cards.

[0121] It should be noted that, in other embodiments, the indicator logic expansion module can be used to execute any step in the data processing method, the indicator logic encapsulation module can be used to execute any step in the data processing method, the indicator interpretability analysis module can be used to execute any step in the data processing method, and the indicator analysis result processing and interaction effect decomposition module can be used to execute any step in the data processing method. The steps implemented by the indicator logic expansion module, indicator logic encapsulation module, indicator interpretability analysis module, and indicator analysis result processing and interaction effect decomposition module can be specified as needed. By implementing different steps in the data processing method through the indicator logic expansion module, indicator logic encapsulation module, indicator interpretability analysis module, and indicator analysis result processing and interaction effect decomposition module, the full functionality of the data processing device can be achieved.

[0122] This application also provides a computing device 500. For example... Figure 5 As shown, the computing device 500 includes a bus 502, a processor 504, a memory 506, and a communication interface 508. The processor 504, the memory 506, and the communication interface 508 communicate with each other via the bus 502. The computing device 500 can be a server or a terminal device. It should be understood that this application does not limit the number of processors and memories in the computing device 500.

[0123] Bus 502 can be a Peripheral Component Interconnect Express (PCIe) bus, an Extended Industry Standard Architecture (EISA) bus, a Unified Bus (Ubus or UB), a Compute Express Link (CXL) bus, a Cache Coherent Interconnect for Accelerators (CCIX) bus, etc. The Unified Bus is also known as the Lingqu Bus. Buses can be divided into address buses, data buses, control buses, etc. For ease of representation, Figure 5 The bus 504 is represented by only one line, but this does not mean that there is only one bus or one type of bus. The bus 504 may include a path for transmitting information between various components of the computing device 500 (e.g., memory 506, processor 504, communication interface 508). The unified bus may also be referred to as the Lingqu bus.

[0124] Processor 504 may include any one or more computing devices such as a central processing unit (CPU), graphics processing unit (GPU), microprocessor (MP) or digital signal processor (DSP), ASIC, FPGA, CPLD, NPU, SoC, offload card, accelerator card, etc.

[0125] Memory 506 may include volatile memory, such as random access memory (RAM). Processor 504 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD). Furthermore, memory 506 may also be implemented using storage class memory (SCM), phase change memory (PCM), or other types of storage media.

[0126] It is worth noting that the same type of storage medium can be configured in the same computing device to realize the function of memory 506, or two or more types of storage media can be configured to realize the function of memory 506. This application does not limit this.

[0127] The memory 506 stores executable program code, which the processor 504 executes to implement the functions of the aforementioned indicator logic expansion module, indicator logic encapsulation module, indicator interpretability analysis module, indicator analysis result processing module, and interaction effect decomposition module, thereby realizing the data processing method. In other words, the memory 506 stores instructions for executing the data processing method.

[0128] The communication interface 508 uses transceiver modules, such as, but not limited to, network interface cards and transceivers, to enable communication between the computing device 500 and other devices or communication networks.

[0129] This application also provides a computing device cluster. The computing device cluster includes at least one computing device. The computing device can be a server, such as a central server, an edge server, or a local server in a local data center. In some embodiments, the computing device can also be a terminal device such as a desktop computer, a laptop computer, or a smartphone.

[0130] like Figure 6 As shown, the computing device cluster includes at least one computing device 500. The memory 506 of one or more computing devices 500 in the computing device cluster may store the same instructions for performing data processing methods.

[0131] In some possible implementations, the memory 506 of one or more computing devices 500 in the computing device cluster may also store partial instructions for executing data processing methods. In other words, a combination of one or more computing devices 500 can jointly execute instructions for executing data processing methods.

[0132] It should be noted that the memory 506 in different computing devices 500 within the computing device cluster can store different instructions, each used to execute a portion of the functions of the data processing device. That is, the instructions stored in the memory 506 of different computing devices 500 can implement the functions of one or more modules among the indicator logic expansion module, indicator logic encapsulation module, indicator interpretability analysis module, indicator analysis result processing module, and interaction effect decomposition module.

[0133] In some possible implementations, one or more computing devices in a computing device cluster can be connected via a network. This network can be a wide area network (WAN) or a local area network (LAN), etc. Figure 7 One possible implementation is shown. For example... Figure 7 As shown, the two computing devices 500A and 500B are connected via a network. Specifically, they are connected to the network through communication interfaces in each computing device. In this possible implementation, the memory 506 in computing device 500A stores instructions for executing the functions of the indicator logic expansion module. Simultaneously, the memory 506 in computing device 500B stores instructions for executing the functions of the indicator logic encapsulation module, the indicator interpretability analysis module, the indicator analysis result processing module, and the interaction effect decomposition module.

[0134] It should be understood that Figure 7 The functions of computing device 500A shown can also be performed by multiple computing devices 500. Similarly, the functions of computing device 500B can also be performed by multiple computing devices 500.

[0135] This application also provides another computing device cluster. The connection relationships between the computing devices in this computing device cluster can be similarly referred to... Figure 6 and Figure 7 The connection method of the computing device cluster is different in that the memory 506 of one or more computing devices 500 in the computing device cluster can store the same instructions for executing data processing methods.

[0136] In some possible implementations, the memory 506 of one or more computing devices 500 in the computing device cluster may also store partial instructions for executing data processing methods. In other words, a combination of one or more computing devices 500 can jointly execute instructions for executing data processing methods.

[0137] This application also provides a computer program product containing instructions. The computer program product may be a software or program product containing instructions, capable of running on a computing device or stored on any usable medium. When the computer program product is run on at least one computing device, it causes the at least one computing device to perform the method provided in this application.

[0138] This application also provides a computer program product containing instructions. The computer program product may be a software or program product containing instructions, capable of running on a computing device or stored on any usable medium. When the computer program product is run on at least one computing device, it causes the at least one computing device to perform the method provided in this application.

[0139] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data processing method, characterized in that, include: Multiple data indicators are acquired, including a first data indicator and multiple second data indicators. The multiple second data indicators are influencing factors that affect the change of the first value, and the first value is the value of the first data indicator. An interpretable logical model is obtained based on the logical relationship between the first data indicator and the plurality of second data indicators. The interpretable logical model is used to obtain the first value based on the plurality of second values, where the second values ​​are the values ​​of the second data indicators. Based on the first value and the plurality of second values, a preset interpretable algorithm is used to calculate the contribution value of the second data indicator in the interpretable logic model. The contribution value is used to quantify the degree of influence of the second data indicator on the numerical change of the first data indicator.

2. The method according to claim 1, characterized in that, The step of calculating the contribution value of the second data indicator in the interpretable logic model based on the first value and the plurality of second values ​​using a preset interpretable algorithm includes: Based on the first value and the plurality of second values, the first contribution value of the second data indicator is obtained by using the Shapley Additivity Model Interpretation Algorithm (SHAP). The first contribution value is obtained based on the Shapley value of the second data indicator. Based on the first value and the plurality of second values, the second contribution value of the second data index is obtained by using an integral gradient algorithm. The second contribution value is obtained based on the integral gradient value of the second data index.

3. The method according to claim 2, characterized in that, After calculating the contribution value of the second data indicator in the interpretable logic model using a preset interpretable algorithm based on the first value and the plurality of second values, the method further includes: Based on multiple first contribution values ​​and multiple second contribution values, an attribution result for the numerical change of the first data indicator is determined, wherein the attribution result includes multiple contribution values ​​of the multiple second data indicators.

4. The method according to claim 3, characterized in that, The step of determining the attribution result of the numerical change of the first data indicator based on multiple first contribution values ​​and multiple second contribution values ​​includes: When the first contribution value of each of the second data indicators is consistent with the second contribution value, the multiple first contribution values ​​of the multiple second data indicators or the multiple second contribution values ​​of the multiple second data indicators are used as the attribution result; When the first contribution value of the second data indicator is inconsistent with the second contribution value, the attribution result is determined based on a combination of multiple data indicators. Each combination of multiple data indicators includes multiple third data indicators, which are obtained by combining the multiple second data indicators.

5. The method according to claim 4, characterized in that, Determining the attribution result based on a combination of multiple data indicators includes: The SHAP algorithm is used to calculate multiple third contribution values ​​for each of the multiple data indicator combinations, and the third contribution values ​​are obtained based on the shapley value of the third data indicator. The integral gradient algorithm is used to calculate multiple fourth contribution values ​​for each of the multiple data indicator combinations, and the fourth contribution values ​​are obtained based on the integral gradient value of the third data indicator. Based on the plurality of third contribution values ​​and the plurality of fourth contribution values, a target data indicator combination is determined from the plurality of data indicator combinations, wherein the third contribution value and the fourth contribution value of each data indicator in the target data indicator combination are equal and each data indicator is independent of the others; The attribution result is determined based on the combination of the target data indicators.

6. The method according to claim 5, characterized in that, Determining the attribution result based on the target data indicator combination includes: The attribution result is determined based on the third or fourth contribution value of multiple target data indicators in the target data indicator combination.

7. A data processing apparatus, characterized in that, include: The indicator logic expansion module is used to obtain multiple data indicators, including a first data indicator and multiple second data indicators. The multiple second data indicators are influencing factors that affect the change of the first value, and the first value is the value of the first data indicator. The indicator logic encapsulation module is used to obtain an interpretable logic model based on the logical relationship between the first data indicator and the plurality of second data indicators. The interpretable logic model is used to obtain the first value based on the plurality of second values, where the second values ​​are the values ​​of the second data indicators. The indicator interpretability analysis module is used to calculate the contribution value of the second data indicator in the interpretable logic model based on the first value and the plurality of second values ​​using a preset interpretable algorithm. The contribution value is used to quantify the degree of influence of the second data indicator on the numerical change of the first data indicator.

8. The apparatus according to claim 7, characterized in that, The indicator interpretation and analysis module is specifically used for: Based on the first value and the plurality of second values, the first contribution value of the second data indicator is obtained by using the Shapley Additivity Model Interpretation Algorithm (SHAP). The first contribution value is obtained based on the Shapley value of the second data indicator. Based on the first value and the plurality of second values, the second contribution value of the second data index in the interpretable logic model is obtained by using the integral gradient algorithm. The second contribution value is obtained based on the integral gradient value of the second data index.

9. The apparatus according to claim 8, characterized in that, After obtaining the multiple contribution values ​​of the multiple second data indicators in the interpretable logic model using an interpretable algorithm, the device further includes: The indicator analysis result processing and interaction effect decomposition module is used to determine the attribution result of the numerical change of the first data indicator based on multiple first contribution values ​​and multiple second contribution values, wherein the attribution result includes multiple contribution values ​​of the multiple second data indicators.

10. The apparatus according to claim 9, characterized in that, The module for processing the index analysis results and decomposing the interaction effect is specifically used for: When the first contribution value of each of the second data indicators is consistent with the second contribution value, the multiple first contribution values ​​of the multiple second data indicators or the multiple second contribution values ​​of the multiple second data indicators are used as the attribution result; When the first contribution value of the second data indicator is inconsistent with the second contribution value, the attribution result is determined based on a combination of multiple data indicators. Each combination of multiple data indicators includes multiple third data indicators, which are obtained by combining the multiple second data indicators.

11. The apparatus according to claim 10, characterized in that, The module for processing the index analysis results and decomposing the interaction effect is specifically used for: The SHAP algorithm is used to calculate multiple third contribution values ​​for each of the multiple data indicator combinations, and the third contribution values ​​are obtained based on the shapley value of the third data indicator. The integral gradient algorithm is used to calculate multiple fourth contribution values ​​for each of the multiple data indicator combinations, and the fourth contribution values ​​are obtained based on the integral gradient value of the third data indicator. Based on the plurality of third contribution values ​​and the plurality of fourth contribution values, a target data indicator combination is determined from the plurality of data indicator combinations, wherein the third contribution value and the fourth contribution value of each data indicator in the target data indicator combination are equal and each data indicator is independent of the others; The attribution result is determined based on the combination of the target data indicators.

12. The apparatus according to claim 11, characterized in that, The module for processing the index analysis results and decomposing the interaction effect is specifically used for: The attribution result is determined based on the third or fourth contribution value of multiple target data indicators in the target data indicator combination.

13. A computing device, characterized in that, The computing device includes a processor and memory; The processor is configured to execute instructions stored in the memory to cause the computing device to perform the method as described in any one of claims 1 to 6.

14. A computing device cluster, characterized in that, It includes at least one computing device, said at least one computing device including a processor and a memory; The processor of the at least one computing device is configured to execute instructions stored in the memory of the at least one computing device to cause the cluster of computing devices to perform the operational steps of the method as described in any one of claims 1 to 6.

15. A computer-readable storage medium, characterized in that, It includes computer program instructions, which, when executed by a cluster of computing devices, perform the operational steps of the method as described in any one of claims 1 to 6.

16. A computer program product containing instructions, characterized in that, When the instruction is executed by the computing device cluster, the computing device cluster causes the computing device cluster to perform the operation steps of the method as described in any one of claims 1 to 6.

Images