Index transaction attribution analysis method and device, computer equipment and storage medium

Abstract

The invention discloses an index transaction attribution analysis method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining problem information, and decomposing the problem information into an intention, a business index and an index dimension; determining whether the user question needs to be subjected to index attribution or not; if the index attribution needs to be carried out, determining whether the business index is abnormal or not; if the business index is abnormal, determining an attribution diagnosis model according to the intention and the index type, and predicting an attribution result of the business index through the attribution diagnosis model; and determining a target external factor and a target internal factor which influence the attribution result, and obtaining countermeasures of the target internal factor or the target external factor. Through dialogue questioning and introduction of a large language model and an attribution analysis model, the reason of abnormal fluctuation of the index can be quickly positioned, minute-level attribution analysis is realized through a dialogue mode, and the method has wide application scenarios and potential.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to an indicator change attribution analysis method, device, computer equipment and storage medium. Background Art

[0002] Currently, data query and business analysis mainly rely on dashboards and business analysis platforms, which require a large number of data analysts to calculate business indicators and integrate them into the platform before presenting them to users through dashboards. For simple queries, users need to manually analyze dashboard tables to obtain data, and indicator fluctuation attribution diagnosis also relies on manual dimensional analysis of contribution.

[0003] However, this model has problems: first, the attribution efficiency is low. The speed and efficiency of manual operation under complex calculation rules are far less than that of the model, which consumes a lot of time and labor; second, attribution is prone to errors. When manually processing large amounts of data, it is easy to make mistakes due to limited attention and concentration, and manual analysis is very limited; third, the analysis model is single and can only attribute the current indicators in multiple dimensions; fourth, the analysis scenario is incomplete and can only meet the attribution analysis of single indicators. Composite indicators need to trace the upstream indicator analysis, which the current technical solution cannot meet; fifth, the specific reasons and countermeasures that affect the changes in indicators are not given based on actual internal and external factors. There are certain problems in attribution efficiency, real-time, comprehensiveness, reducing labor costs, and improving decision-making efficiency. Summary of the invention

[0004] Based on this, it is necessary to provide an indicator change attribution analysis method, device, computer equipment and storage medium to address the above technical issues, so as to solve at least one problem existing in the above-mentioned prior art.

[0005] In a first aspect, an embodiment of the present application provides an indicator change attribution analysis method, including: Obtain problem information, and decompose the problem information into intent, business indicators, and indicator dimensions; Determine whether the user problem requires metric attribution; If indicator attribution is required, determine whether the business indicator is abnormal; If the business indicator is abnormal, determine an attribution diagnosis model according to the intention and indicator dimension, so as to predict the attribution result of the business indicator through the attribution diagnosis model; Determine the target external factors and target internal factors that affect the attribution result, and obtain the response measures for the target internal factors or the target external factors.

[0006] In one embodiment, determining whether the business indicator is abnormal includes: Determine whether there is an abnormality in the business indicator at the current time; and / or Determine whether the numerical difference of the business data corresponding to the business indicator in different comparison periods is greater than a preset difference threshold.

[0007] In one embodiment, determining whether there is an abnormality in the business indicator at the current time includes: Determine the mean of the business data corresponding to all business indicators within each preset time range; Determine whether the difference between the mean and the mean of the business data corresponding to all business indicators within an adjacent preset time range is greater than a preset difference; If so, the business indicator is abnormal; If not, determine the change trend of the business data corresponding to each indicator dimension; When the difference between the peak value and the trough value of the change trend is greater than a preset threshold, the business indicator is abnormal.

[0008] In one embodiment, the attribution diagnosis model is a single-dimensional diagnosis model, and the attribution result of the business indicator is predicted by the attribution diagnosis model, including: Calculate the contribution value of the business data corresponding to each indicator dimension respectively; The contribution values ​​are sorted in ascending or descending order to obtain the indicator dimension with the highest contribution value as the attribution result of the business indicator.

[0009] In one embodiment, the attribution diagnosis model is a composite indicator blood relationship diagnosis model, and the attribution result of the business indicator is predicted by the attribution diagnosis model, including: When the business indicator is a composite indicator, query the upstream indicator of the business indicator to obtain multiple parent indicators of the business indicator; Calculate the impact value of each parent indicator; The impact values ​​are sorted in ascending or descending order to obtain the parent indicator with the highest impact value as the attribution result of the business indicator.

[0010] In one embodiment, the attribution diagnosis model is a dimensional cross-attribution diagnosis model, and the attribution result of the business indicator is predicted by the attribution diagnosis model, including: Combine the indicator dimensions to obtain multiple groups of multidimensional indicator dimensions; Calculate the contribution value of each indicator dimension under multiple groups of multi-dimensional indicator dimensions; A set of indicator dimensions with the largest contribution value is taken as the attribution result of the business indicator.

[0011] In one embodiment, problem information is obtained and decomposed into intent, business indicators and indicator dimensions, including: Setting prompt words, wherein the prompt words include dimension information and indicator metadata; Acquire the problem information, and determine the indicator to be attributed according to the problem information; Querying the database for attribution dimension metadata corresponding to the indicator to be attributed; Based on the prompt words, question information and the attribution dimension metadata, semantic analysis is performed through a large language model to obtain a DSL object; The database is queried according to the DSL object to obtain indicator data.

[0012] In a second aspect, an indicator change attribution analysis device is provided, comprising: A problem information decomposition unit, used to obtain problem information and decompose the problem information into intentions, business indicators and indicator dimensions; An indicator attribution determination unit, used to determine whether the user problem needs indicator attribution; An indicator abnormality analysis unit, used to determine whether the business indicator is abnormal if indicator attribution is required; an attribution result prediction unit, configured to determine an attribution diagnosis model according to the intention and the indicator dimension if the business indicator is abnormal, so as to predict the attribution result of the business indicator through the attribution diagnosis model; The countermeasure determination unit is used to determine the target external factors and target internal factors that affect the attribution result, and obtain the countermeasures of the target internal factors or the target external factors.

[0013] In a third aspect, a computer device is provided, comprising a memory, a processor, and computer-readable instructions stored in the memory and running on the processor, wherein the processor implements the indicator change attribution analysis method as described above when executing the computer-readable instructions.

[0014] In a fourth aspect, a readable storage medium is provided, on which computer-readable instructions are stored. When the computer-readable instructions are executed by a processor, the indicator change attribution analysis method as described above is implemented.

[0015] The above-mentioned indicator change attribution analysis method, device, computer equipment and storage medium, and its method implementation include: obtaining problem information, decomposing the problem information into intent, business indicators and indicator dimensions; determining whether the user problem needs to be attributed to indicators; if indicator attribution is required, determining whether the business indicator is abnormal; if the business indicator is abnormal, determining an attribution diagnosis model based on the intent and indicator dimension, so as to predict the attribution result of the business indicator through the attribution diagnosis model; determining the target external factors and target internal factors that affect the attribution result, and obtaining the response measures of the target internal factors or the target external factors. In the embodiment of the present application, through dialogue questioning, introducing a large language model and an attribution analysis model, the cause of abnormal fluctuations of indicators can be quickly located, and minute-level attribution analysis can be realized through dialogue. Minute-level attribution analysis is realized, and by querying the influencing factor knowledge base, internal and external factors that affect indicator fluctuations are automatically identified, providing a specific basis for adjusting operational strategies, which has a wide range of application scenarios and potential. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments of the present application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0017] Figure 1 This is a flow chart of an indicator change attribution analysis method in one embodiment of the present application; Figure 2 This is a structural schematic diagram of an indicator change attribution analysis device in one embodiment of the present application; Figure 3 It is a schematic diagram of a computer device in one embodiment of the present application. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0019] In one embodiment, if Figure 1 As shown, a method for attribution analysis of indicator changes is provided, comprising the following steps: In step S110, problem information is obtained, and the problem information is decomposed into intention, business indicator and indicator dimension; In an embodiment of the present application, the questions input by the user can be decomposed by a large language model, such as GPT4o. Specifically, a prompt template can be set in advance. When the question information is obtained, the to-be-attributed indicator corresponding to the question information can be determined, and the attribution dimension metadata corresponding to the to-be-attributed indicator can be queried. Then, a semantic analysis request is generated through the prompt, the attribution dimension metadata, and the question information, and a DSL object is obtained based on the precise semantic analysis of the GPT4o. A corresponding attribution analysis request is generated according to the DSL object, and the database is queried to obtain the indicator data.

[0020] In step S120, it is determined whether the user question needs to be attributed to an indicator; In the embodiment of the present application, the understanding ability and rules of the large language model can be called, for example, configurable keywords such as fluctuation, abnormality, attribution analysis, abnormality, etc., to determine whether indicator attribution is needed. For example, if the question is "Why did the xxx indicator drop so much", the large language model can be used to identify abnormal fluctuations in the indicator and whether attribution analysis is needed.

[0021] In step S130, if indicator attribution is required, it is determined whether the business indicator is abnormal; In an embodiment of the present application, anomaly identification can be performed on the current date corresponding to the indicator. Anomaly identification can be performed from two aspects. First, the indicator can be used to see whether the overall trend is abnormal. For example, the average of the business data corresponding to each indicator within a preset time range can be calculated, such as the daily average. Then, based on the average, a line graph is established. According to the overall trend of the lines in the line graph, it is determined whether there are large fluctuations. Second, it can be determined whether the business indicator data corresponding to each dimension is abnormal. A line graph can be established for the indicator data of each dimension, and then whether the trend of the line fluctuates significantly. If there is a large fluctuation, it means that the business indicator is abnormal.

[0022] In the embodiment of the present application, a comparative analysis can be performed on the specified period of the indicator to determine whether there is an abnormality, such as day-on-day, week-on-week, month-on-month, year-on-year, etc., or other specified time, holidays, etc., and the corresponding values ​​and difference values ​​of the indicator data in two different periods are calculated. If the difference value is greater than the preset difference threshold, it indicates that there is an abnormality.

[0023] It should be noted that the current date corresponding to the indicator can be identified for anomalies. If there is no anomaly, a comparative analysis can be performed on the specified period of the indicator. Alternatively, a comparative analysis can be performed on the specified period of the indicator first. If it is determined that there is no anomaly in the indicator, the current date corresponding to the indicator can be identified for anomalies again. Alternatively, the current date corresponding to the indicator can be identified for anomalies and the specified period of the indicator can be compared and analyzed at the same time. Alternatively, either one can be selected to determine the anomaly of the business indicator.

[0024] In step S140, if the business indicator is abnormal, an attribution diagnosis model is determined according to the intention and the indicator dimension, so as to predict the attribution result of the business indicator through the attribution diagnosis model; Among them, the attribution diagnosis model may include a single-dimensional diagnosis model, a composite indicator bloodline diagnosis model, and a dimensional cross-attribution diagnosis model. All indicators can be directly diagnosed through a single-dimensional diagnosis model. If the indicator is a composite indicator, it can be accurately diagnosed through a composite indicator bloodline diagnosis model.

[0025] When a business indicator is abnormal, one or more attribution diagnosis models can be determined according to the indicator attribution intention and the indicator type, and the attribution diagnosis model can be used to predict the attribution result of the business indicator. For example, when the business indicator is a composite indicator, the composite indicator attribution lineage diagnosis model can be selected to predict the attribution result.

[0026] In step S150, target external factors and target internal factors that affect the attribution result are determined, and countermeasures for the target internal factors or the target external factors are obtained.

[0027] In the embodiments of the present application, internal factors refer to internal operating strategies, technological changes and other factors. The operating strategy may specifically include distributing red envelopes / driver subsidies (which will affect the number of user orders and the driver's order acceptance index), price reductions (driver response rate, order acceptance rate index) and other operational activities. The technological changes refer to new product features (which will affect the user driver experience index) and technical problems (which will affect the user driver usage index, order placement and acceptance index, etc.).

[0028] Among them, external factors refer to road conditions, weather, traffic and other factors.

[0029] In an embodiment of the present application, the business database can be queried to obtain internal factors such as operating strategies, activities, and technology release records related to the indicator dimension to determine whether they are attribution results that affect business indicators. For example, a corresponding impact level can be configured for each internal factor, and then a corresponding weight can be assigned to each internal factor based on the impact level. The higher the impact level, the greater the weight and the higher the impact coefficient. Similarly, external factors can also be configured with corresponding impact levels and assigned weights. Based on this, the impact of internal and external factors on the attribution results can be determined. It should be noted that an attribution routing engine can be set to associate indicators with internal and external factors. When the indicators fluctuate, the weights of internal and external factors can be determined, thereby determining the internal and external factors that ultimately affect the attribution results.

[0030] In an embodiment of the present application, external factors, such as weather, such as rain, snow, etc., will affect the supply and demand indicators between drivers and consumers. Road conditions, such as congestion and traffic restrictions, will affect the actual average order completion time indicator. For policies, such as holding events, affecting driver distribution indicators, etc. Therefore, different response measures can be implemented for different external factors. For example, for weather, the order price can usually be increased to attract drivers to accept orders. In response to road conditions, transportation capacity can be dispatched to allow drivers in other areas to accept orders. In response to policies, order acceptance subsidies can be provided to dispatch vehicles to other areas with fewer vehicles.

[0031] In the embodiments of the present application, internal factors include operating strategies, product changes, etc. Operating strategies, such as bonus distribution and subsidies, will affect consumer order indicators and driver order acceptance indicators. Price cuts will affect driver response rates and order acceptance rate indicators. Product changes, such as new product features, will affect consumer driver experience indicators. Technical failures will affect consumer and driver usage indicators. Therefore, in response to the above internal factors, product functions may be managed, or technical failure repairs, emergency expansion and contraction, and other measures may be taken.

[0032] In an embodiment of the present application, an indicator change attribution analysis method is provided, including: obtaining problem information, decomposing the problem information into intent, business indicators and indicator dimensions; determining whether the user problem needs to be attributed to indicators; if indicator attribution is required, determining whether the business indicator is abnormal; if the business indicator is abnormal, determining an attribution diagnosis model based on the intent and indicator dimensions, so as to predict the attribution result of the business indicator through the attribution diagnosis model; determining the target external factors and target internal factors that affect the attribution result, and obtaining the response measures of the target internal factors or the target external factors. In an embodiment of the present application, through dialogue questioning, introducing a large language model and an attribution analysis model, the cause of abnormal fluctuations in indicators can be quickly located, and minute-level attribution analysis can be achieved. Minute-level attribution analysis is achieved through dialogue, and by querying the influencing factor knowledge base, internal and external factors that affect indicator fluctuations are automatically identified, providing a specific basis for adjusting operational strategies, which has a wide range of application scenarios and potential.

[0033] In an embodiment of the present application, determining whether the business indicator is abnormal includes: Determine whether there is an abnormality in the business indicator at the current time; and / or Determine whether the numerical difference of the business data corresponding to the business indicator in different comparison periods is greater than a preset difference threshold.

[0034] Optionally, to determine whether the business indicator at the current time is abnormal, abnormality identification can be performed from two aspects. First, the overall trend can be checked through the indicator to see whether there is an abnormality. For example, the average of the business data corresponding to each indicator within a preset time range, such as the daily average, can be calculated, and then a line graph can be established based on the average to see whether there is a large fluctuation in the overall trend of the lines in the line graph. Second, it can be determined whether the business indicator data corresponding to each dimension is abnormal, that is, a line graph can be established for the indicator data of each dimension, and then whether the trend of the line fluctuates significantly. If there is a large fluctuation, it means that the business indicator is abnormal.

[0035] Determine whether the numerical difference of the business data corresponding to the business indicator in different comparison periods is greater than the preset difference threshold, such as daily year-on-year, weekly year-on-year, monthly year-on-year, year-on-year, etc., or other specified time, holidays, etc., and calculate the corresponding numerical value and difference value of the indicator data in two different periods. If the difference value is greater than the preset difference threshold, it means that there is an abnormality. It should be noted that the corresponding value can be queried in the doris library table according to the problem information, the identified indicator and the date. For example: to calculate the weekly year-on-year comparison of the order completion quantity indicator, the indicator value of the order completion quantity today and 7 days ago can be queried in the doris library, and then the difference value = (current value-benchmark value) / benchmark value*100%. Among them, the current value refers to the indicator value in the current period, and the benchmark value refers to the indicator value in the same period last year.

[0036] In an embodiment of the present application, determining whether there is an abnormality in the business indicator at the current time includes: Determine the mean of the business data corresponding to all business indicators within each preset time range; Determine whether the difference between the mean and the mean of the business data corresponding to all business indicators within an adjacent preset time range is greater than a preset difference; If so, the business indicator is abnormal; If not, determine the change trend of the business data corresponding to each indicator dimension; When the difference between the peak value and the trough value of the change trend is greater than a preset threshold, the business indicator is abnormal.

[0037] Optionally, the mean of the business data corresponding to all business indicators within each preset time range can be calculated, such as the mean of business data within each day, that is, the daily average, and then a line graph can be established to determine whether there are large fluctuations based on the overall trend of the lines in the line graph. It can be determined whether the difference between different daily averages is large. If so, it can indicate that the business data is abnormal. If not, the value of the business data corresponding to each indicator dimension can be further calculated, and a line graph can be established. Based on the changing trend of the lines, it can be determined whether there are large fluctuations, that is, when the difference between the peak value and the trough value of the line is greater than the preset threshold, if so, it can indicate that the business indicator is abnormal.

[0038] In an embodiment of the present application, the attribution diagnosis model is a single-dimensional diagnosis model, and the attribution result of the business indicator is predicted by the attribution diagnosis model, including: Calculate the contribution value of the business data corresponding to each indicator dimension respectively; The contribution values ​​are sorted in ascending or descending order to obtain the indicator dimension with the highest contribution value as the attribution result of the business indicator.

[0039] Optionally, the business data can be grouped according to the indicator dimension, and then the contribution value of the business data under each indicator dimension is calculated. The contribution value can be calculated by a paired weight fluctuation algorithm, or can also be obtained by a fluctuation contribution algorithm. According to the calculated contribution value of each indicator dimension, it can be arranged in ascending or descending order, and then the maximum contribution value is used as the result of this attribution analysis.

[0040] Among them, the pairing weight fluctuation algorithm is specifically as follows: the pairing order weight of each dimension can be set, for example, the proportion of pairing orders on the benchmark day of dimension m / (1-the proportion of pairing orders on the benchmark day of dimension m), then the contribution value = (dimension m change rate-total change rate)*dimension pairing order weight.

[0041] The volatility contribution algorithm is as follows: The volatility contribution algorithm of the absolute value indicator: the change in dimension m divided by the total benchmark value. The volatility contribution algorithm of the rate value indicator: (the current rate value of dimension m - the overall benchmark rate value) * (the absolute value of the denominator of the current rate value / the overall absolute value of the denominator of the current rate value) - (the benchmark rate value of dimension m - the overall benchmark rate value) * (the absolute value of the denominator of the benchmark rate value / the overall absolute value of the denominator of the benchmark rate value). The current rate value of dimension m refers to the ratio or proportion of dimension m starting from this week, and the current rate value refers to a certain ratio or proportion within the current cycle or time period.

[0042] In one embodiment of the present application, the attribution diagnosis model is a composite indicator blood relationship diagnosis model, and the attribution result of the business indicator is predicted by the attribution diagnosis model, including: When the business indicator is a composite indicator, query the upstream indicator of the business indicator to obtain multiple parent indicators of the business indicator; Calculate the impact value of each parent indicator; The impact values ​​are sorted in ascending or descending order to obtain the parent indicator with the highest impact value as the attribution result of the business indicator.

[0043] Optionally, after obtaining the business indicator, it can be determined whether it is a composite indicator. If so, the indicator's blood relationship is queried in the indicator system, that is, the indicator's upstream indicator, for example, if the customer unit price = total order amount / total order volume, the parent indicator of the customer unit price can be the two indicators of total order amount and total order volume. Then, the impact value can be calculated by the chain substitution method, and then the impact value of each parent indicator can be sorted in ascending or descending order to obtain the parent indicator with the highest impact value as the attribution result of the business indicator.

[0044] The calculation process of the chain substitution method is as follows: Assume that a certain analysis index M is obtained by three interrelated factors A, B, and C. The current index and the comparative index are: Current indicator M1=A1 * B1 * C1 Comparison index M0=A0 * B0 * C0 When determining the degree of influence of each factor change on the index, it is necessary to sort them in a certain order (the importance is from high to low, and the order of arrangement will affect the final result): Comparison index M0 = A0 * B0 * C0……(1) First substitution A1 * B0 * C0……(2) Second substitution A1 * B1 * C0……(3) The third substitution is A1 * B1 * C1……(4) (2)-(1)→The impact of changes in A on M.

[0045] (3)-(2)→The impact of changes in B on M.

[0046] (4)-(3)→The impact of changes in C on M.

[0047] Total impact: △M = M1 - M0 = (4)-(3)+(3)-(2)+(2)-(1).

[0048] For example, if sales amount = number of exposures * conversion rate * average order value decreases year-on-year, the impact of each indicator can be determined through the above chain substitution method. The details are shown in the following table: Table 1: Calculation results of the chain substitution algorithm

[0049] Based on the above Table 1, we can see that the final exposure quantity affects 22,400, the conversion rate affects -9,600, and the average order value affects 14,400.

[0050] In an embodiment of the present application, the attribution diagnosis model is a dimensional cross-attribution diagnosis model, and the attribution result of the business indicator is predicted by the attribution diagnosis model, including: Combine the indicator dimensions to obtain multiple groups of multidimensional indicator dimensions; Calculate the contribution value of each indicator dimension under multiple groups of multi-dimensional indicator dimensions; A set of indicator dimensions with the largest contribution value is taken as the attribution result of the business indicator.

[0051] Optionally, the business data can be grouped according to the indicator dimension, and then the contribution value of each indicator dimension can be calculated. For example, if there are 3 dimensions, after combining them in pairs, 6 multi-dimensional indicator dimension combinations can be obtained, and the contribution value of the business data under each indicator dimension can be calculated twice.

[0052] The contribution value can be calculated through a paired weight fluctuation algorithm, or can also be obtained through a fluctuation contribution algorithm. According to the calculated contribution value of each indicator dimension, they can be arranged in ascending or descending order, and then a group of indicator dimensions with the largest contribution value can be used as the result of this attribution analysis.

[0053] In an embodiment of the present application, problem information is obtained and decomposed into intent, business indicators and indicator dimensions, including: Setting prompt words, wherein the prompt words include dimension information and indicator metadata; Acquire the problem information, and determine the indicator to be attributed according to the problem information; Querying the database for attribution dimension metadata corresponding to the indicator to be attributed; Based on the prompt words, question information and the attribution dimension metadata, semantic analysis is performed through a large language model to obtain a DSL object; The database is queried according to the DSL object to obtain indicator data.

[0054] Optionally, a prompt template (prompt word) can be set in advance, and the prompt contains information such as the dimensions and indicator metadata of the business data. When the problem information is obtained, the to-be-attributed indicator corresponding to the problem information can be determined, and then the attribution dimension metadata corresponding to the to-be-attributed indicator can be queried in the database. Then, based on the problem information, prompt template, and attribution dimension metadata, a semantic analysis request is generated, and the semantic analysis request is processed through the large language model GPT4o to obtain a DSL object, and then the corresponding attribution analysis request is generated based on the DSL object, and the query SQL is generated to query the database to obtain the indicator data.

[0055] For example, a user asks "Why has the order volume for freight trucks in region A decreased compared to last week?" By utilizing the information extraction capability of the big model, we can obtain the parsing result of the big model in Json format. The big model identifies dimensions (city / business line / vehicle type, etc.), indicators, and intentions based on the prompt. The dimensions are: City: Region A, Business line: freight truck, Indicator to be attributed: Order volume, Current time: Time: 2024-11-18, Same period time: Time: 2024-11-11, Intention: Attribute the reason for the decline.

[0056] In the embodiment of the present application, by introducing a large language model and an attribution diagnosis model through a dialogue question, the cause of abnormal fluctuations in indicators can be quickly located, and minute-level attribution analysis can be achieved. The results of single-dimensional diagnosis, composite indicator blood diagnosis, multi-dimensional cross-diagnosis and other models can improve the accuracy and efficiency of attribution analysis. Through the knowledge base, the external and internal factors that affect the fluctuation of indicators are queried, and corresponding measures are obtained for processing, providing effective decision-making suggestions. Due to its high efficiency and accuracy, the present application can be adapted to a variety of scenarios, such as: Enterprise operation monitoring: In the daily operation of an enterprise, the fluctuations of various key indicators (such as sales, number of users, conversion rate, etc.) are crucial to operational decisions. Through the above scheme, the abnormal fluctuations of these indicators can be quickly identified, and the attribution analysis can be automatically given to help the operation team quickly adjust the strategy and ensure the stable operation of the enterprise. Financial risk management: In the financial field, changes in market risk, credit risk, etc. often lead to fluctuations in various financial indicators. Through the above scheme, financial institutions can monitor the abnormal changes of these indicators in real time, quickly identify potential risks, and take corresponding risk management measures. E-commerce platform analysis: Indicators such as product sales and user behavior on e-commerce platforms are crucial for merchants. The above solutions can help merchants quickly identify the reasons for declining sales, user loss and other problems, so as to adjust marketing strategies and improve user experience and sales. Medical health monitoring: In the field of medical health, fluctuations in various physiological indicators of patients (such as heart rate, blood pressure, blood sugar, etc.) are crucial for doctors' diagnosis and treatment. Although the above solutions are mainly used for attribution analysis of non-medical data, their concepts and technologies can also be used in medical health monitoring. By real-time monitoring of patients' physiological indicators, abnormal fluctuations can be quickly identified to provide doctors with timely diagnosis and treatment suggestions. Supply chain management: In supply chain management, indicators such as inventory turnover rate and order processing time are crucial to the operational efficiency of enterprises. Through the above solutions, enterprises can quickly identify bottlenecks and problems in the supply chain, thereby optimizing the supply chain management process and improving operational efficiency. In these scenarios, the implementation of the above solutions can help professionals in related fields quickly understand the reasons behind data changes, so as to make more accurate and timely decisions.

[0057] It should be understood that the size of the serial numbers of the steps in the above embodiments 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 application.

[0058] In one embodiment, a device for analyzing the attribution of index changes is provided, and the device for analyzing the attribution of index changes corresponds to the method for analyzing the attribution of index changes in the above embodiment. Figure 2As shown, the indicator change attribution analysis device includes a problem information decomposition unit 10, an indicator attribution determination unit 20, an indicator abnormality analysis unit 30, an attribution result prediction unit 40, and a countermeasure determination unit 50. The functional modules are described in detail as follows: A problem information decomposition unit 10 is used to obtain problem information and decompose the problem information into intentions, business indicators and indicator dimensions; An indicator attribution determination unit 20, used to determine whether the user question needs indicator attribution; An indicator abnormality analysis unit 30, used to determine whether the business indicator is abnormal if indicator attribution is required; The attribution result prediction unit 40 is used to determine an attribution diagnosis model according to the intention and the indicator dimension if the business indicator is abnormal, so as to predict the attribution result of the business indicator through the attribution diagnosis model; The countermeasure determination unit 50 is used to determine the target external factors and target internal factors that affect the attribution result, and obtain the countermeasures of the target internal factors or the target external factors.

[0059] In one embodiment of the present application, the indicator abnormality analysis unit 30 is further used to: Determine whether there is an abnormality in the business indicator at the current time; and / or Determine whether the numerical difference of the business data corresponding to the business indicator in different comparison periods is greater than a preset difference threshold.

[0060] In one embodiment of the present application, the indicator abnormality analysis unit 30 is further used to: Determine the mean of the business data corresponding to all business indicators within each preset time range; Determine whether the difference between the mean and the mean of the business data corresponding to all business indicators within an adjacent preset time range is greater than a preset difference; If so, the business indicator is abnormal; If not, determine the change trend of the business data corresponding to each indicator dimension; When the difference between the peak value and the trough value of the change trend is greater than a preset threshold, the business indicator is abnormal.

[0061] In an embodiment of the present application, the attribution diagnosis model is a single-dimensional diagnosis model, and the attribution result prediction unit 40 is further used to: The attribution diagnosis model is a single-dimensional diagnosis model. The attribution result of the business indicator is predicted by the attribution diagnosis model, including: Calculate the contribution value of the business data corresponding to each indicator dimension respectively; The contribution values ​​are sorted in ascending or descending order to obtain the indicator dimension with the highest contribution value as the attribution result of the business indicator.

[0062] In one embodiment of the present application, the attribution diagnosis model is a composite indicator blood relationship diagnosis model, and the attribution result prediction unit 40 is further used to: When the business indicator is a composite indicator, query the upstream indicator of the business indicator to obtain multiple parent indicators of the business indicator; Calculate the impact value of each parent indicator; The impact values ​​are sorted in ascending or descending order to obtain the parent indicator with the highest impact value as the attribution result of the business indicator.

[0063] In an embodiment of the present application, the attribution diagnosis model is a dimensional cross-attribution diagnosis model, and the attribution result prediction unit 40 is further used to: Combine the indicator dimensions to obtain multiple groups of multidimensional indicator dimensions; Calculate the contribution value of each indicator dimension under multiple groups of multi-dimensional indicator dimensions; A set of indicator dimensions with the largest contribution value is taken as the attribution result of the business indicator.

[0064] In one embodiment of the present application, the question information decomposition unit 10 is further used to: Setting prompt words, wherein the prompt words include dimension information and indicator metadata; Acquire the problem information, and determine the indicator to be attributed according to the problem information; Querying the database for attribution dimension metadata corresponding to the indicator to be attributed; Based on the prompt words, question information and the attribution dimension metadata, semantic analysis is performed through a large language model to obtain a DSL object; The database is queried according to the DSL object to obtain indicator data.

[0065] In the embodiment of the present application, by introducing a large language model and an attribution diagnosis model through dialogue questioning, the cause of abnormal fluctuations in indicators can be quickly located, and minute-level attribution analysis can be achieved. The results of single-dimensional diagnosis, composite indicator bloodline diagnosis, multi-dimensional cross-diagnosis and other models can improve the accuracy and efficiency of attribution analysis. Through the knowledge base, the external and internal factors that affect the fluctuation of indicators are queried, and corresponding measures are obtained for processing, providing effective decision-making suggestions. Due to its high efficiency and accuracy, this application can be adapted to a variety of scenarios.

[0066] For the specific limitations of the indicator change attribution analysis device, please refer to the limitations of the indicator change attribution analysis method above, which will not be repeated here. Each module in the above-mentioned indicator change attribution analysis device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0067] In one embodiment, a computer device is provided. The computer device may be a terminal device, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a readable storage medium. The readable storage medium stores computer-readable instructions. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer-readable instructions are executed by the processor, an indicator change attribution analysis method is implemented. The readable storage medium provided in this embodiment includes a non-volatile readable storage medium and a volatile readable storage medium.

[0068] In an embodiment of the present application, a computer device is provided, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, wherein the processor implements the steps of the indicator change attribution analysis method as described above when executing the computer-readable instructions.

[0069] In an embodiment of the application, a readable storage medium is provided, wherein the readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the indicator change attribution analysis method as described above are implemented.

[0070] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing related hardware through computer-readable instructions, and the computer-readable instructions can be stored in a non-volatile readable storage medium or a volatile readable storage medium. When the computer-readable instructions are executed, they may include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0071] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0072] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for attribution analysis of indicator changes, characterized in that: The method comprises: Obtain problem information, and decompose the problem information into intent, business indicators, and indicator dimensions; Determine whether the user problem requires metric attribution; If indicator attribution is required, determine whether the business indicator is abnormal; If the business indicator is abnormal, determine an attribution diagnosis model according to the intention and indicator dimension, so as to predict the attribution result of the business indicator through the attribution diagnosis model; Determine the target external factors and target internal factors that affect the attribution result, and obtain the response measures for the target internal factors or the target external factors.

2. The indicator change attribution analysis method according to claim 1, characterized in that: The determining whether the business indicator is abnormal includes: Determine whether there is an abnormality in the business indicator at the current time; and / or Determine whether the numerical difference of the business data corresponding to the business indicator in different comparison periods is greater than a preset difference threshold.

3. The indicator change attribution analysis method according to claim 2, characterized in that: Determining whether there is an abnormality in the business indicator at the current time includes: Determine the mean of the business data corresponding to all business indicators within each preset time range; Determine whether the difference between the mean and the mean of the business data corresponding to all business indicators within an adjacent preset time range is greater than a preset difference; If so, the business indicator is abnormal; If not, determine the change trend of the business data corresponding to each indicator dimension; When the difference between the peak value and the trough value of the change trend is greater than a preset threshold, the business indicator is abnormal.

4. The indicator change attribution analysis method according to claim 1, characterized in that: The attribution diagnosis model is a single-dimensional diagnosis model. The attribution result of the business indicator is predicted by the attribution diagnosis model, including: Calculate the contribution value of the business data corresponding to each indicator dimension respectively; The contribution values ​​are sorted in ascending or descending order to obtain the indicator dimension with the highest contribution value as the attribution result of the business indicator.

5. The indicator change attribution analysis method according to claim 1, characterized in that: The attribution diagnosis model is a composite indicator blood relationship diagnosis model. The attribution result of the business indicator is predicted by the attribution diagnosis model, including: When the business indicator is a composite indicator, query the upstream indicator of the business indicator to obtain multiple parent indicators of the business indicator; Calculate the impact value of each parent indicator; The impact values ​​are sorted in ascending or descending order to obtain the parent indicator with the highest impact value as the attribution result of the business indicator.

6. The indicator change attribution analysis method according to claim 1, characterized in that: The attribution diagnosis model is a dimensional cross-attribution diagnosis model. The attribution result of the business indicator is predicted by the attribution diagnosis model, including: Combine the indicator dimensions to obtain multiple groups of multidimensional indicator dimensions; Calculate the contribution value of each indicator dimension under multiple groups of multi-dimensional indicator dimensions; A set of indicator dimensions with the largest contribution value is taken as the attribution result of the business indicator.

7. The indicator change attribution analysis method according to claim 1, characterized in that: The obtaining of problem information, decomposing the problem information into intention, business indicator and indicator dimension, includes: Setting prompt words, wherein the prompt words include dimension information and indicator metadata; Acquire the problem information, and determine the indicator to be attributed according to the problem information; Querying the database for attribution dimension metadata corresponding to the indicator to be attributed; Based on the prompt words, question information and attribute dimension metadata, semantic analysis is performed through a large language model to obtain a DSL object; The database is queried according to the DSL object to obtain indicator data.

8. An indicator change attribution analysis device, characterized in that: The device comprises: A problem information decomposition unit, used to obtain problem information and decompose the problem information into intentions, business indicators and indicator dimensions; An indicator attribution determination unit, used to determine whether the user problem needs indicator attribution; An indicator abnormality analysis unit, used to determine whether the business indicator is abnormal if indicator attribution is required; an attribution result prediction unit, configured to determine an attribution diagnosis model according to the intention and the indicator dimension if the business indicator is abnormal, so as to predict the attribution result of the business indicator through the attribution diagnosis model; The countermeasure determination unit is used to determine the target external factors and target internal factors that affect the attribution result, and obtain the countermeasures of the target internal factors or the target external factors.

9. A computer device comprising a memory, a processor, and computer-readable instructions stored in the memory and executed on the processor, characterized in that: When the processor executes the computer-readable instructions, the indicator change attribution analysis method as described in any one of claims 1 to 7 is implemented.

10. A readable storage medium having computer readable instructions stored thereon, characterized in that: When the computer-readable instructions are executed by a processor, the indicator change attribution analysis method as described in any one of claims 1 to 7 is implemented.

Images