Index difference attribution method based on index hierarchical disassembly

Abstract

The invention provides an index difference attribution method based on index hierarchical disassembly, which comprises the following steps of: firstly carrying out hierarchical disassembly on attribution indexes, then carrying out attribution fusion calculation, calculating contribution rates of the attribution indexes on different index disassembly layers, and carrying out weighted product calculation on each hierarchical result to obtain an attribution result; not only can the contribution rate of the attribution index on different index disassembly layers be visually analyzed and checked, but also the overall contribution rate of the attribution index can be obtained. The logic of index hierarchical disassembly is more in line with the general logic of enterprise business index disassembly, and the mode of calculating the depth index contribution rate by the weighted product provides a simple and visual index contribution transverse comparison mode for business users. According to the embodiment of the invention, the contribution rate of each layer is calculated through service data splitting, when the attribution index changes, the disassembly index can be updated and calculated uniformly, and the service analysis efficiency can be improved while the calculation pressure is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of data set processing, and in particular to an indicator gap attribution method based on indicator hierarchical decomposition, an indicator gap attribution device based on indicator hierarchical decomposition, an electronic device and a computer-readable storage medium. Background Art

[0002] Business operations usually involve the analysis of indicators of business data, and the business situation is analyzed by analyzing the changes in data under different indicators.

[0003] Generally, business analysis of enterprises will analyze the core indicators (also called attribution indicators) from different indicator paths. In view of the hierarchical structure of retail sales, the overall retail sales need to be analyzed from three levels of indicators. The indicators are divided into addition and subtraction and multiplication and division, and different decomposition schemes may be used at each level. For example, the analysis of the retail sales of a footwear and clothing retail enterprise will first be decomposed into the retail sales of old customers and the retail sales of new customers (first level), which is an addition decomposition; the retail sales of old customers will be further decomposed into the unit price of old customers × the number of orders of old customers (second level), which is a multiplication decomposition; the unit price of old customers will be decomposed into the product of the average price of old customers' tags, the discount rate of old customers, and the joint rate of old customers, which is also a multiplication decomposition (third level).

[0004] Therefore, in the indicator analysis at different levels, whether it is addition or multiplication of indicator operations, the calculation results reflect the contribution of each layer and each calculation indicator to the core indicator (overall retail amount) (the contribution rate corresponding to different indicators).

[0005] However, in market operations, some core indicators may change. For example, if the overall retail sales amount changes, the retail sales amount breakdown indicators will also change. At this time, if the attribution indicators are calculated once (only on the first level), it will undoubtedly increase the calculation pressure and the workload of subsequent indicator analysis, which is inefficient.

[0006] In addition, when the overall retail sales amount changes, how to intuitively evaluate the contribution of each indicator to the overall retail sales amount? According to the above-mentioned method of analyzing and calculating each disassembled indicator, at present, we can only gradually understand and evaluate the contribution of each indicator to the overall retail sales amount based on each disassembled indicator. Therefore, it is impossible to intuitively evaluate the contribution of each indicator to the overall retail sales amount, and the evaluation is time-consuming. The calculation and analysis of traditional operating indicators is mainly single-layer calculation, which cannot visualize the contribution of attribution indicators at subsequent levels, and it is impossible to intuitively view the influence of attribution indicators at each level. Summary of the invention

[0007] In order to solve the technical problems existing in the prior art, the present invention provides the following technical solutions: On the one hand, a method for attributing an indicator gap based on hierarchical decomposition of indicators is provided, the method being implemented by an electronic device, and the method comprising: S1. Determine the indicator decomposition path of the attribution indicator; S2. Perform layered decomposition of contribution rate along the indicator decomposition path; S3, importing the business data of the attribution indicator, and calculating the contribution rate of the attribution indicator at different indicator decomposition layers in a hierarchical manner; S4. Perform weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition levels to obtain the overall contribution rate of the attribution indicators.

[0008] Preferably, S1, determining an indicator decomposition path of the attribution indicator, includes: Determine the business analysis scope of the attribution metric; Extracting keywords of the attribution indicator; Based on the keywords and the business analysis scope, construct prompt words and input them into a preset LLM large language model; Automatically mining and outputting the attribution indicator analysis logic associated with the prompt word through the LLM large language model; Reading the logical nodes and logical subjects in the attribution indicator analysis logic; Divide the indicator decomposition layer according to the logical nodes; Using the logic node in the indicator decomposition layer as the decomposition indicator of the attribution indicator in the indicator decomposition layer; The indicator decomposition path of the attribution indicator is determined according to the determined indicator decomposition layer and the decomposition indicator.

[0009] Preferably, S2, performing layered decomposition of contribution rate along the indicator decomposition path, comprises: The attribution indicators are additively decomposed or multiplicatively decomposed at different indicator decomposition layers, wherein: The addition is broken down into: Let A be the attribution indicator and ai be the decomposition indicator, then: , , The multiplication is broken down into: Let B be the attribution index and bi be the decomposition index, then: , , , The superscripts 1 and 0 represent the current period and the base period respectively.

[0010] Preferably, S3, importing the business data of the attribution indicator, and calculating the contribution rate of the attribution indicator at different indicator decomposition layers in a hierarchical manner, includes: Importing business data of the attribution indicator; Decomposing the business data according to the decomposition indicators of the attribution indicators at different indicator decomposition layers to generate business sub-data corresponding to the decomposition indicators; According to the business sub-data, the contribution rates of the attribution indicators at different indicator decomposition layers are subjected to addition decomposition calculation or multiplication decomposition calculation, and the contribution rates of the attribution indicators at different indicator decomposition layers are obtained by layered calculation.

[0011] Preferably, S4, performing weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition layers to obtain the overall contribution rate of the attribution indicators, includes: Read the contribution rates of the attribution indicators at different indicator decomposition levels, which are recorded as α1, α2 and α3 respectively; The contribution rates of the attribution index at different index decomposition levels: α1, α2 and α3 are weighted product calculations to obtain the overall contribution rate α of the attribution index: α=α1×α2×α3×100%.

[0012] Preferably, the method further comprises: Calculate the absolute contribution value β of the attribution indicator: β=C×α, Wherein, C represents the business quota of the comparison period preset for the attribution indicator, that is, the business quota of the attribution indicator in a certain period is used as the comparison parameter.

[0013] On the other hand, an indicator gap attribution device based on indicator hierarchical disassembly is provided, and the indicator gap attribution device based on indicator hierarchical disassembly is used to implement the indicator gap attribution method based on indicator hierarchical disassembly, and the device includes: The path confirmation module is used to determine the indicator decomposition path of the attribution indicator; A contribution rate hierarchical decomposition module, used for performing hierarchical decomposition of the contribution rate along the indicator decomposition path; A hierarchical calculation module, used for importing the business data of the attribution indicator, and hierarchically calculating the contribution rate of the attribution indicator at different indicator decomposition layers; The attribution calculation module is used to perform weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition layers to obtain the overall contribution rate of the attribution indicators.

[0014] On the other hand, an electronic device is provided, comprising: a processor; a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, any one of the above-mentioned indicator gap attribution methods based on indicator hierarchical decomposition is implemented.

[0015] On the other hand, a computer-readable storage medium is provided, in which at least one instruction is stored. The at least one instruction is loaded and executed by a processor to implement any one of the above-mentioned indicator gap attribution methods based on indicator hierarchical decomposition.

[0016] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: The present invention determines the indicator decomposition path of the attribution indicator; performs hierarchical decomposition of the contribution rate along the indicator decomposition path; imports the business data of the attribution indicator, and hierarchically calculates the contribution rate of the attribution indicator at different indicator decomposition layers; performs weighted product calculation on the contribution rate of the attribution indicator at different indicator decomposition layers to obtain the overall contribution rate of the attribution indicator. The attribution indicator can be first decomposed into layers and then the attribution fusion calculation can be performed, the contribution rate of the attribution indicator at different indicator decomposition layers can be calculated, and the weighted product calculation of each layered result can be performed. Not only can the contribution rate of the attribution indicator at different indicator decomposition layers be visualized and viewed, but also the overall contribution rate of the attribution indicator can be obtained. The logic of hierarchical decomposition of indicators is more in line with the general logic of enterprise business indicator decomposition. The method of calculating the contribution rate of deep indicators by weighted product provides business users with a simple and intuitive method for horizontal comparison of indicator contributions.

[0017] By splitting business data and calculating the contribution rate of each layer, the decomposed indicators can be uniformly updated and calculated when the attribution indicators change, which can improve business analysis efficiency while reducing computing pressure. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 This is a schematic diagram of a business indicator analysis of an indicator hierarchical analysis provided by an embodiment of the present invention; Figure 2 It is a flow chart of an indicator gap attribution method based on indicator hierarchical decomposition provided by an embodiment of the present invention; Figure 3It is a visual analysis diagram of a hierarchical decomposition analysis of attribution indicators provided by an embodiment of the present invention; Figure 4 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0021] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0022] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same. "of", "corresponding, relevant" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same.

[0023] In the embodiments of the present invention, sometimes the subscripts such as W 1 It may be mistakenly written as a non-subscript form such as W1. When the difference is not emphasized, the meanings they express are the same.

[0024] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0025] The embodiment of the present invention provides an indicator gap attribution method based on indicator hierarchical decomposition, which can be implemented by an electronic device, which can be a terminal or a server. Figure 2 The flowchart of the indicator gap attribution method based on indicator hierarchical decomposition is shown in FIG. The processing flow of the method may include the following steps: S1. Determine the indicator decomposition path of the attribution indicator; S2. Perform layered decomposition of contribution rate along the indicator decomposition path; S3, importing the business data of the attribution indicator, and calculating the contribution rate of the attribution indicator at different indicator decomposition layers in a hierarchical manner; S4. Perform weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition levels to obtain the overall contribution rate of the attribution indicators.

[0026] The attribution index (core index) of this embodiment can be determined according to business attributes, such as retail sales amount.

[0027] In this embodiment, for the decomposition indicators of attribution indicators at different business analysis layers (such as Figure 1 For example, the decomposition indicators at the first level include the retail amount of old customers and the retail amount of new customers. The other decomposition indicators at the second and third levels are as follows: Figure 1 As shown in the figure), it can be determined according to the business analysis level or by the administrator.

[0028] Each step will be described in detail below.

[0029] Preferably, S1, determining an indicator decomposition path of the attribution indicator, includes: Determine the business analysis scope of the attribution metric; Extracting keywords of the attribution indicator; Based on the keywords and the business analysis scope, construct prompt words and input them into a preset LLM large language model; Automatically mining and outputting the attribution indicator analysis logic associated with the prompt word through the LLM large language model; Reading the logical nodes and logical subjects in the attribution indicator analysis logic; Divide the indicator decomposition layer according to the logical nodes; Using the logic node in the indicator decomposition layer as the decomposition indicator of the attribution indicator in the indicator decomposition layer; The indicator decomposition path of the attribution indicator is determined according to the determined indicator decomposition layer and the decomposition indicator.

[0030] First, it is necessary to determine the indicator decomposition path of the attribution indicator, that is, the business analysis level of the attribution indicator, and from which business analysis logic layers the contribution rate of the attribution indicator needs to be analyzed.

[0031] In order to quickly analyze the attribution analysis logic of the attribution indicators and realize intelligent and automated attribution logic analysis, this embodiment uses the LLM large language model (a self-developed model, or a third-party large language model such as Baidu's Wenxin Yiyan) to assist in confirming the indicator decomposition path of the attribution indicators.

[0032] The analysis logic of attribution indicators in business analysis mainly refers to what kind of business analysis process should be followed and which decomposed indicators should be analyzed during business analysis. Therefore, as long as the logical nodes and logical subjects in the analysis logic are determined, the indicator decomposition path of the attribution indicator can be confirmed. The indicator decomposition path is directional and should be analyzed in a directional manner according to the analysis logic of the attribution indicator.

[0033] Therefore, the present invention allows the administrator to confirm the business analysis scope of the attribution indicator in advance, and then combine the keywords of the attribution indicator (such as the analysis process of the retail amount in the XX business analysis, the keywords may include "retail, amount, xx business", etc.) to construct the prompt words of the large language model, and then input the prompt words into the large language model, so that the large language model can automatically mine and generate the corresponding attribution indicator analysis logic based on the prompt words. The large language model can automatically analyze the keywords and business analysis scope in the prompt words, and generate the attribution indicator analysis logic of the corresponding business logic (similar to the current LLM (such as Baidu Wenxin Yiyan) technology that automatically generates corresponding copy / methods and steps based on prompt words). Subsequently, the analysis logic is identified with logical nodes, and the attribution indicators are divided into several layers of indicator logic analysis layers according to the logical nodes.

[0034] In each logical analysis layer, you can also confirm the logical subjects in different analysis layers according to the logical description in the analysis logic. For example, the analysis logic of retail amount under the indicator of old customer retail amount is the first-level old customer retail amount. The logical subject includes old customer retail finance, and the attribution indicator will calculate the contribution rate of old customer retail amount in the first level. After analyzing the old customer retail amount, enter the next level of order analysis.

[0035] The specific steps of LLM to automatically generate attribution indicator analysis logic based on prompt words may include: ‌Prompt word analysis‌: ‌Input processing‌: Receive prompt words entered by the user. These prompt words may be natural language descriptions, such as "Analysis and calculation of retail amounts in orders for xx brand shoes."

[0036] ‌Semantic Analysis‌: Use natural language processing (NLP) techniques, such as word segmentation, part-of-speech tagging, and syntactic analysis, to understand the intent and structure of the prompt word.

[0037] ‌Intent Recognition‌: ‌Determine business scenarios‌: Based on the parsed prompt words, identify specific business scenarios, such as user registration, order processing, data query, etc.

[0038] Extract key information: Extract key business logic elements from the prompt words, such as trigger conditions, operation objects, execution actions, etc.

[0039] ‌Logical Construction‌: ‌Rule definition‌: Based on the results of intent recognition and key information extraction, define specific business rules. These rules may include conditional judgment, data operation, process control, etc.

[0040] ‌Logic diagram generation‌: In some cases, you can generate visual diagrams of business logic, such as flow charts, decision trees, etc., to more intuitively understand and modify business logic.

[0041] ‌Code Generation‌: Template matching: Generates corresponding business logic code based on predefined code templates and logic construction results. These codes may include functions, classes, methods, etc. in programming languages.

[0042] ‌Custom Code‌: For logic that cannot be completely generated by template matching, you may need to write custom code to implement specific functionality.

[0043] ‌Verification and Testing‌: ‌Logic Verification‌: Use unit testing, integration testing and other methods to verify whether the generated business logic meets the expected behavior and results.

[0044] ‌Performance Optimization‌: Evaluate and optimize the performance of the generated code to ensure that it can run efficiently in the actual environment.

[0045] ‌Deployment and Integration‌: ‌System Integration‌: Integrate the generated business logic code into existing business systems, such as databases, application servers, etc.

[0046] ‌Deployment and launch‌: After testing and optimization, the business logic is deployed to the production environment for user use.

[0047] ‌Monitoring and Maintenance‌: ‌Operation Monitoring‌: Real-time monitoring of the deployed business logic to ensure its normal operation and detect potential problems in a timely manner.

[0048] Please implement it in combination with the principles and applications of LLM.

[0049] Preferably, S2, performing layered decomposition of contribution rate along the indicator decomposition path, comprises: The attribution indicators are additively decomposed or multiplicatively decomposed at different indicator decomposition layers, wherein: The addition is broken down into: Let A be the attribution indicator and ai be the decomposition indicator, then: , ; The multiplication is broken down into: Let B be the attribution index and bi be the decomposition index, then: , , , The superscripts 1 and 0 represent the current period and the base period respectively.

[0050] The additive and multiplicative decomposition of attribution indicators can be combined with the attached Figure 1 The disassembly layers are schematically described.

[0051] For example, the retail amount performs additive or multiplicative decomposition on the attribution indicator at the first level of additive decomposition, where: The addition is broken down into: , The contribution rate of retail sales of old customers to the change in retail sales + the contribution rate of retail sales of new customers to the change in retail sales = 100%.

[0052] For example, the multiplication of retail sales at the first level is as follows: , The contribution rate of the average order value of old customers to the change of the retail amount of old customers + the contribution rate of the number of orders from old customers to the change of the retail amount of old customers = 100%.

[0053] For other similar indicators, they can be broken down and calculated in a similar way. For example, the average customer price can be broken down into discount rate, joint rate, average tag price, and so on.

[0054] Preferably, S3, importing the business data of the attribution indicator, and calculating the contribution rate of the attribution indicator at different indicator decomposition layers in a hierarchical manner, includes: Importing business data of the attribution indicator; Decomposing the business data according to the decomposition indicators of the attribution indicators at different indicator decomposition layers to generate business sub-data corresponding to the decomposition indicators; According to the business sub-data, the contribution rates of the attribution indicators at different indicator decomposition layers are subjected to addition decomposition calculation or multiplication decomposition calculation, and the contribution rates of the attribution indicators at different indicator decomposition layers are obtained by layered calculation.

[0055] The business data of the attribution indicator can be counted from the front end and uploaded to the back end at a regular time every day, and the back end will record and bind it to the corresponding attribution indicator.

[0056] After the background receives the business data of the attribution indicator, it first performs data analysis on the business data of the attribution indicator according to the preset indicator decomposition path to obtain the business sub-data of the attribution indicator at different indicator decomposition layers, and the business sub-data needs to be configured as the business sub-data under the corresponding decomposition indicators at each layer.

[0057] Combined with Figure 3 As shown, by parsing the business data and writing the corresponding decomposition indicators, the business data of the attribution indicators are filled under the decomposition indicators of the corresponding analysis layer, allowing the administrator to visually view the business sub-data of the attribution indicators under the decomposition indicators at each level, which is convenient for the administrator to perform visual analysis on the attribution indicators. Through visual analysis, it is possible to clearly see the data differences of the attribution indicators at different decomposition layers, and clearly understand the indicator data and contribution rate of the attribution indicators under each decomposition indicator, providing business users with a simple and intuitive horizontal comparison method of indicator contributions. Compared with traditional business background display board technology, the present invention is more capable of visually analyzing the decomposition indicator data of the attribution indicators in different business layers and the contribution rate of the corresponding decomposition indicators to the attribution indicators, so as to find the indicator differences that affect the attribution indicators, allowing users to locate and analyze the indicator data with low contribution rates and adjust business positioning.

[0058] According to the business sub-data, the contribution rate of the attribution indicator at different indicator decomposition layers (analysis layers) is calculated by addition or multiplication, and the contribution rate of the attribution indicator at different indicator decomposition layers is obtained by layered calculation. Therefore, the indicator decomposition calculation method of the attribution indicator at different analysis layers is specified above. Therefore, after the decomposed indicator is written into the corresponding business sub-data, the contribution rate can be calculated according to the indicator decomposition calculation method configured at the layer. Finally, the contribution rate of the attribution indicator at different analysis layers is calculated by weighted product to obtain the overall contribution rate of the attribution indicator.

[0059] Preferably, S4, performing weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition layers to obtain the overall contribution rate of the attribution indicators, includes: Read the contribution rates of the attribution indicators at different indicator decomposition levels, which are recorded as α1, α2 and α3 respectively; The contribution rates of the attribution index at different index decomposition levels: α1, α2 and α3 are weighted product calculations to obtain the overall contribution rate α of the attribution index: α=α1×α2×α3×100%.

[0060] The contribution rate to the overall indicator is obtained by multiplying the contribution rate indicators on the indicator decomposition path. For example: .

[0061] The contribution rate of other attribution indicators to the overall indicators can refer to the above method.

[0062] Preferably, the method further comprises: Calculate the absolute contribution value β of the attribution indicator: β=C×α, Wherein, C represents the business quota of the comparison period preset for the attribution indicator, that is, the business quota of the attribution indicator in a certain period is used as the comparison parameter.

[0063] After obtaining the overall contribution rate, you can also evaluate the expected value of the attribution indicator. You can calculate the expected data of the attribution indicator (such as expected retail amount, sales) by multiplying the retail amount in the comparison period (set by the administrator, or select a set of retail amounts from historical data as the retail amount in the comparison period) with the overall contribution rate of the attribution indicator. For example: The overall contribution of the old customer discount rate to the retail amount = the retail amount in the comparison period × the overall contribution rate of the old customer discount rate.

[0064] Therefore, the overall contribution rate can be combined to make expected predictions of the attribution indicators.

[0065] S4, performing weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition levels to obtain the overall contribution rate of the attribution indicators, further comprising: Counting the business data of the attribution indicator and the overall contribution rate α of the attribution indicator, and generating a contribution rate log of the attribution indicator; Constructing a contribution rate log data set consisting of contribution rate logs of several attribution indicators; Based on deep learning technology, feature engineering and model training are performed on the contribution rate log data set to generate an attribution indicator contribution rate prediction model; The attribution indicator contribution rate prediction model is deployed on a background server and used to predict the overall contribution rate of the attribution indicator.

[0066] In addition to the above applications, in order to quickly calculate the overall contribution rate of the attribution indicator in the corresponding business, the present embodiment can also perform statistics on the business data of the attribution indicator calculated each time and the overall contribution rate α of the attribution indicator, so as to construct a contribution rate log data set composed of the contribution rate logs of several of the attribution indicators.

[0067] The data set contains a large number of attribution indicator contribution rate statistical parameters, which reflect the contribution rates made by different attribution indicators based on different business data.

[0068] Therefore, the present invention can use deep learning technology to predict the overall contribution rate of attribution indicators under different business data, so that administrators can know in advance the possible contribution rate and possible expected value of the attribution indicators (further predict and calculate the absolute contribution value based on the predicted overall contribution rate).

[0069] Therefore, we can extract features from the business data and the corresponding contribution rates in the data set, extract features and build a feature set. Then, we divide the data into training sets and validation sets according to the proportion, and import the training sets into the preset deep learning model (such as random forest or decision tree) for feature classification learning, so as to train and obtain the contribution rates that can identify different business data features.

[0070] The specific deep learning model used, such as RNN or random forest model, is not described in detail in this embodiment and is determined by the administrator. The following is a training scheme for a random forest model: Using Random Forest for data feature classification training usually includes the following steps: 1. Data Preparation a. Data Collection Collect a data set containing the target variable (i.e. the variable to be predicted) and the feature variable (i.e. the variable used for prediction). Here are the business data and overall contribution rates of different attribution indicators.

[0071] b. Data preprocessing ‌Missing Value Handling‌: Fill in or remove missing values.

[0072] ‌Data cleaning‌: remove outliers, duplicate values, etc.

[0073] Feature Encoding: Convert categorical features to numerical features, such as using One-Hot Encoding or Label Encoding.

[0074] ‌Feature Scaling‌: Standardize or normalize numerical features as needed, but this is usually not necessary in random forests because they are insensitive to feature scale.

[0075] 2. Data Partitioning Divide the data set into a training set and a validation set, usually using 70%-80% of the data as the training set and the remaining data as the validation set. You can also use cross-validation to more comprehensively evaluate model performance.

[0076] 3. Build a random forest model a. Import necessary libraries In Python, the scikit-learn library is often used to build random forest models.

[0077] b. Instantiate the random forest classifier Set the parameters of the random forest model, such as the number of trees, maximum depth, etc.

[0078] c. Training the model Train a random forest model using the training set data.

[0079] 4. Model Evaluation a. Prediction validation set Use the trained random forest model to make predictions on the validation set.

[0080] b. Evaluate performance Calculate the model's accuracy, recall, F1 score and other metrics, and draw the confusion matrix to evaluate the model performance.

[0081] The specific test is completed by the administrator.

[0082] 5. Model tuning (optional) According to the model evaluation results, adjust the parameters of the random forest, such as increasing or decreasing the number of trees, limiting the maximum depth of the tree, adjusting the minimum number of sample splits, etc., to optimize the model performance.

[0083] 6. Feature Importance Analysis Random forests provide feature importance scores that can help us understand which features contribute most to the model's predictions.

[0084] 7. Model Saving and Deployment When the model performance meets the requirements, you can save the model as a file so that it can be loaded and used in subsequent applications. In Python, you can use the joblib or pickle library to save and load the model.

[0085] After the model is trained, it can be deployed on the background to predict the contribution rate of attribution indicators.

[0086] The above are the basic steps for using random forests to perform data feature classification training. Depending on the specific data set and task requirements, these steps may need to be appropriately adjusted and optimized.

[0087] Another aspect of the present invention provides a block diagram of an indicator gap attribution device based on indicator hierarchical disassembly, and the device is used in an indicator gap attribution method based on indicator hierarchical disassembly. The device includes: The path confirmation module is used to determine the indicator decomposition path of the attribution indicator; A contribution rate hierarchical decomposition module, used for performing hierarchical decomposition of the contribution rate along the indicator decomposition path; A hierarchical calculation module, used for importing the business data of the attribution indicator, and hierarchically calculating the contribution rate of the attribution indicator at different indicator decomposition layers; The attribution calculation module is used to perform weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition layers to obtain the overall contribution rate of the attribution indicators.

[0088] The functions and interactions of the above modules can be connected in accordance with the principles in the above method steps, which will not be described in detail here.

[0089] Figure 4 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Figure 4 As shown, the electronic device may include the above Figure 3 The indicator gap attribution device based on indicator hierarchical decomposition is shown. Optionally, the electronic device 410 may include a first processor 2001 .

[0090] Optionally, the electronic device 410 may further include a memory 2002 and a transceiver 2003 .

[0091] The first processor 2001, the memory 2002 and the transceiver 2003 may be connected via a communication bus.

[0092] Combine the following Figure 4 The components of the electronic device 410 are described in detail: The first processor 2001 is the control center of the electronic device 410, and may be a processor or a general term for multiple processing elements. For example, the first processor 2001 is one or more central processing units (CPUs), or may be application specific integrated circuits (ASICs), or may be one or more integrated circuits configured to implement the embodiments of the present invention, such as one or more microprocessors (digital signal processors, DSPs), or one or more field programmable gate arrays (field programmable gate arrays, FPGAs).

[0093] Optionally, the first processor 2001 can execute various functions of the electronic device 410 by running or executing a software program stored in the memory 2002 and calling data stored in the memory 2002 .

[0094] In a specific implementation, as an embodiment, the first processor 2001 may include one or more CPUs, such as Figure 4 CPU0 and CPU1 are shown in FIG.

[0095] In a specific implementation, as an embodiment, the electronic device 410 may also include multiple processors, such as Figure 4The first processor 2001 and the second processor 2004 are shown in FIG. Each of these processors may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). The processor here may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0096] The memory 2002 is used to store the software program for executing the solution of the present invention, and is controlled to be executed by the first processor 2001. The specific implementation method can refer to the above method embodiment, which will not be repeated here.

[0097] Optionally, the memory 2002 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 2002 may be integrated with the first processor 2001, or may exist independently and access the first processor 2001 through the interface circuit ( Figure 4 (not shown) is coupled to the first processor 2001, which is not specifically limited in this embodiment of the present invention.

[0098] The transceiver 2003 is used to communicate with a network device or a terminal device.

[0099] Optionally, the transceiver 2003 may include a receiver and a transmitter ( Figure 4 The receiver is used to implement a receiving function, and the transmitter is used to implement a sending function.

[0100] Optionally, the transceiver 2003 may be integrated with the first processor 2001, or may exist independently and communicate with the first processor 2001 through the interface circuit ( Figure 4 (not shown) is coupled to the first processor 2001, which is not specifically limited in this embodiment of the present invention.

[0101] It should be noted that Figure 4 The structure of the electronic device 410 shown in the figure does not constitute a limitation on the router, and the actual knowledge structure recognition device may include more or fewer components than those shown in the figure, or combine certain components, or arrange the components differently.

[0102] In addition, the technical effects of the electronic device 410 can refer to the technical effects of the indicator gap attribution method based on indicator hierarchical decomposition described in the above method embodiment, and will not be repeated here.

[0103] It should be understood that the first processor 2001 in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0104] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0105] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0106] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0107] In the present invention, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0108] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0109] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0110] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0111] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0112] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0113] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0114] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0115] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for attributing indicator gaps based on hierarchical decomposition of indicators, characterized in that: The method comprises: S1. Determine the indicator decomposition path of the attribution indicator; S2. Perform layered decomposition of contribution rate along the indicator decomposition path; S3, importing the business data of the attribution indicator, and calculating the contribution rate of the attribution indicator at different indicator decomposition layers in a hierarchical manner; S4. Perform weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition levels to obtain the overall contribution rate of the attribution indicators.

2. The indicator gap attribution method based on indicator hierarchical decomposition according to claim 1 is characterized in that: S1. Determine the indicator decomposition path of the attribution indicator, including: Determine the business analysis scope of the attribution indicator; Extracting keywords of the attribution indicator; Based on the keywords and the business analysis scope, construct prompt words and input them into a preset LLM large language model; Automatically mining and outputting the attribution indicator analysis logic associated with the prompt word through the LLM large language model; Reading the logical nodes and logical subjects in the attribution indicator analysis logic; Divide the indicator decomposition layer according to the logical nodes; Using the logic node in the indicator decomposition layer as the decomposition indicator of the attribution indicator in the indicator decomposition layer; The indicator decomposition path of the attribution indicator is determined according to the determined indicator decomposition layer and the decomposition indicator.

3. The indicator gap attribution method based on indicator layering and decomposition according to claim 2 is characterized in that: S2. Perform layered decomposition of contribution rate along the indicator decomposition path, including: The attribution indicators are additively decomposed or multiplicatively decomposed at different indicator decomposition layers, wherein: The addition is broken down into: Let A be the attribution indicator and ai be the decomposition indicator, then: ， ， The multiplication is broken down into: Let B be the attribution index and bi be the decomposition index, then: ， ， ， The superscripts 1 and 0 represent the current period and the base period respectively.

4. The indicator gap attribution method based on indicator layering and decomposition according to claim 3 is characterized in that: S3. Importing the business data of the attribution indicator, and calculating the contribution rate of the attribution indicator at different indicator decomposition layers in a hierarchical manner, including: Importing business data of the attribution indicator; Decomposing the business data according to the decomposition indicators of the attribution indicators at different indicator decomposition layers to generate business sub-data corresponding to the decomposition indicators; According to the business sub-data, the contribution rates of the attribution indicators at different indicator decomposition layers are subjected to addition decomposition calculation or multiplication decomposition calculation, and the contribution rates of the attribution indicators at different indicator decomposition layers are obtained by layered calculation.

5. The indicator gap attribution method based on indicator layering and decomposition according to claim 4 is characterized in that: S4. Perform weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition levels to obtain the overall contribution rate of the attribution indicators, including: Read the contribution rates of the attribution indicators at different indicator decomposition levels, which are recorded as α1, α2 and α3 respectively; The contribution rates of the attribution index at different index decomposition levels: α1, α2 and α3 are weighted product calculations to obtain the overall contribution rate α of the attribution index: 。 6. The indicator gap attribution method based on indicator layering and decomposition according to claim 5 is characterized in that: The method further comprises: Calculate the absolute contribution value β of the attribution indicator: β=C×α， Wherein, C represents the business quota of the comparison period preset for the attribution indicator, that is, the business quota of the attribution indicator in a certain period is used as the comparison parameter.

7. An indicator gap attribution device based on indicator hierarchical disassembly, the indicator gap attribution device based on indicator hierarchical disassembly is used to implement the indicator gap attribution method based on indicator hierarchical disassembly as claimed in any one of claims 1 to 6, characterized in that: The device comprises: The path confirmation module is used to determine the indicator decomposition path of the attribution indicator; A contribution rate hierarchical decomposition module, used for performing hierarchical decomposition of the contribution rate along the indicator decomposition path; A hierarchical calculation module, used for importing the business data of the attribution indicator, and hierarchically calculating the contribution rate of the attribution indicator at different indicator decomposition layers; The attribution calculation module is used to perform weighted product calculation on the contribution rates of the attribution indicators at different indicator decomposition layers to obtain the overall contribution rate of the attribution indicators.

8. An electronic device, characterized in that: The electronic device comprises: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, which can be called by a processor to execute the method according to any one of claims 1 to 6.

Images