Composite index fluctuation attribution method and device, electronic equipment and storage medium

Abstract

The present disclosure relates to a composite index fluctuation attribution method, device, electronic equipment and storage medium. The composite index fluctuation attribution method comprises: determining a composite index corresponding to recommendation information, and obtaining a target factor associated with the recommendation information, the composite index being used to measure the ratio between a display resource corresponding to the recommendation information and an interaction amount; obtaining a first influence degree and a second influence degree of the target factor, wherein the first influence degree reflects the influence of the target factor on the fluctuation of the display resource, and the second influence degree reflects the influence of the target factor on the fluctuation of the interaction amount; and determining a composite index influence degree corresponding to the target factor according to the first influence degree and the second influence degree, wherein the composite index influence degree reflects the influence of the target factor on the fluctuation of the composite index of the recommendation information.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of Internet, and particularly relates to a composite index fluctuation attribution method and device, electronic equipment and storage medium. BACKGROUND

[0002] In the related art, a ratio of a display resource corresponding to recommendation information to an interaction amount is taken as a composite index, which represents how much display resource needs to be put into an information display platform under a certain recommendation effect. When these composite indexes fluctuate greatly, it is necessary to locate the fluctuation source and then make targeted follow-up adjustments. However, the existing fluctuation locating method often relies on manpower to qualitatively analyze possible fluctuation sources, and it is difficult to quantify potential fluctuation sources for composite indexes without additive properties, and the analysis efficiency is low. SUMMARY

[0003] The present disclosure provides a composite index fluctuation attribution method, device, electronic equipment and storage medium to at least solve the problem of low fluctuation analysis efficiency in the related art, and can not solve any of the above problems.

[0004] According to a first aspect of the present disclosure, a composite index fluctuation attribution method is provided, the composite index fluctuation attribution method comprising: determining a composite index corresponding to recommendation information, and obtaining a target factor associated with the recommendation information, the composite index being used to measure a ratio between a display resource corresponding to the recommendation information and an interaction amount; obtaining a first influence degree and a second influence degree of the target factor, wherein the first influence degree reflects an influence of the target factor on a fluctuation of a display resource of the recommendation information, and the second influence degree reflects an influence of the target factor on a fluctuation of an interaction amount of the recommendation information; and determining a composite index influence degree corresponding to the target factor according to the first influence degree and the second influence degree, wherein the composite index influence degree reflects an influence of the target factor on a fluctuation of a composite index of the recommendation information.

[0005] Optionally, the step of determining the composite index influence degree corresponding to the target factor according to the first influence degree and the second influence degree comprises: determining a difference value obtained by subtracting the second influence degree from the first influence degree, and determining the composite index influence degree based on the difference value.

[0006] Optionally, the step of determining the composite index influence degree based on the difference value comprises: taking the difference value as a first composite index influence degree.

[0007] Optionally, the step of determining the composite index influence degree based on the difference value comprises: obtaining an error term of the composite index influence degree; the error term reflects an error of the difference value relative to the composite index influence degree; and adjusting the difference value based on the error term to obtain a second composite index influence degree.

[0008] Optionally, the step of obtaining the error term of the composite index influence degree comprises: determining a product of the first influence degree and the second influence degree as the error term.

[0009] Optionally, the composite index fluctuation attribution method further comprises: selecting at least one attribution factor from the plurality of target factors according to the composite index influence degree, wherein the composite index influence degree of the attribution factor is greater than the composite index influence degrees of other target factors, or the absolute value of the composite index influence degree of the attribution factor is greater than the absolute values of the composite index influence degrees of other target factors.

[0010] Optionally, the step of obtaining the target factor associated with the recommendation information comprises: determining a target dimension associated with the recommendation information; and obtaining the target factor under the target dimension based on the target dimension.

[0011] According to a second aspect of the present disclosure, a composite index fluctuation attribution apparatus is provided, comprising: a first obtaining unit configured to determine a composite index corresponding to recommendation information and obtain a target factor associated with the recommendation information, the composite index being used to measure a ratio between a display resource and an interaction amount corresponding to the recommendation information; a second obtaining unit configured to obtain a first influence degree and a second influence degree of the target factor, wherein the first influence degree reflects an influence of the target factor on a fluctuation of a display resource of the recommendation information, and the second influence degree reflects an influence of the target factor on a fluctuation of an interaction amount of the recommendation information; and a determining unit configured to determine a composite index influence degree corresponding to the target factor according to the first influence degree and the second influence degree, wherein the composite index influence degree reflects an influence of the target factor on a fluctuation of the composite index of the recommendation information.

[0012] Optionally, the determining unit is further configured to determine a difference value obtained by subtracting the second influence degree from the first influence degree, and determine the composite index influence degree based on the difference value.

[0013] Optionally, the determining unit is further configured to take the difference value as a first composite index influence degree.

[0014] Optionally, the determining unit is further configured to: obtain an error term of the composite indicator influence degree; the error term reflects an error of the difference value with respect to the composite indicator influence degree; and adjust the difference value based on the error term to obtain a second composite indicator influence degree.

[0015] Optionally, the determining unit is further configured to: determine a product of the first influence degree and the second influence degree as the error term.

[0016] Optionally, the composite indicator fluctuation attribution apparatus further comprises an attribution unit configured to: select at least one attribution factor from the plurality of target factors according to the composite indicator influence degree, wherein the composite indicator influence degree of the attribution factor is greater than the composite indicator influence degree of other target factors, or the absolute value of the composite indicator influence degree of the attribution factor is greater than the absolute value of the composite indicator influence degree of other target factors.

[0017] Optionally, the first obtaining unit is further configured to: determine a target dimension associated with the recommendation information; and obtain a target factor under the target dimension based on the target dimension.

[0018] According to a third aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; at least one memory storing computer-executable instructions, wherein the computer-executable instructions, when executed by the at least one processor, cause the at least one processor to perform the composite indicator fluctuation attribution method according to the present disclosure.

[0019] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, which, when instructions in the computer-readable storage medium are executed by at least one processor, causes the at least one processor to perform the composite indicator fluctuation attribution method according to the present disclosure.

[0020] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising computer instructions, which, when executed by at least one processor, implement the composite indicator fluctuation attribution method according to the present disclosure.

[0021] The technical solutions provided by the embodiments of the present disclosure at least bring the following beneficial effects:

[0022] The composite index fluctuation attribution method and the composite index fluctuation attribution device according to the embodiments of the present disclosure utilize the additive performance of the display resource and the interaction amount, obtain the first influence degree and the second influence degree of the plurality of target factors respectively, and obtain the composite index influence degree of the corresponding target factor, so that the quantitative attribution of the composite index is converted into the quantitative attribution of the additive index, the quantitative analysis of the composite index fluctuation is possible, and the source causing the composite index fluctuation can be effectively determined from the plurality of target factors. The scheme can realize the quantification of the attribution of the composite index fluctuation, help to improve the analysis efficiency, comprehensively consider the influence of the display resource and the interaction amount, and thus guarantee the comprehensiveness and reliability of the analysis, so that the target factor causing the abnormal change of the composite index can be quickly located.

[0023] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and do not limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0024] The accompanying drawings incorporated in and forming a part of the specification illustrate the embodiments consistent with the present disclosure and serve to explain the principles of the present disclosure, and do not constitute an undue limitation on the present disclosure.

[0025] Figure 1 FIG. 1 is a flowchart illustrating a composite index fluctuation attribution method according to an example embodiment of the present disclosure.

[0026] Figure 2 FIG. 2 is a flowchart illustrating a composite index fluctuation attribution method according to an example embodiment of the present disclosure.

[0027] Figure 3 FIG. 3 is a flowchart illustrating a composite index fluctuation attribution method according to an example embodiment of the present disclosure.

[0028] Figure 4 FIG. 4 is a block diagram illustrating a composite index fluctuation attribution device according to an example embodiment of the present disclosure.

[0029] Figure 5 FIG. 5 is a block diagram illustrating an electronic device according to an example embodiment of the present disclosure. DETAILED DESCRIPTION

[0030] In order for those skilled in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the drawings.

[0031] It should be noted that the terms "first", "second", and the like in the description and in the claims of the present disclosure and above-described accompanying drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0032] It should be noted herein that "at least one of a plurality" appearing in the present disclosure represents three types of alternatives, including "any one of the plurality", "a combination of any multiple of the plurality", and "all of the plurality". For example, "including at least one of A and B" includes the following three alternatives: (1) including A; (2) including B; (3) including A and B. For another example, "performing at least one of step one and step two" represents the following three alternatives: (1) performing step one; (2) performing step two; (3) performing step one and step two.

[0033] In the field of Internet service recommendation, composite indicators such as CPM, CPA (Cost Per Action), CPC (Cost Per Click) are often used to calculate the display resources that the service party (i.e. the party with service recommendation needs) needs to invest in the information display platform. Assuming that the values of the composite indicators are 10 / CPM, 10 / CPA, and 10 / CPC respectively, then correspondingly, the service party needs to invest 10 units of display resources in the information display platform for every 1000 exposures of the recommended information, every execution of the predetermined action by the target user of the service party (i.e. the user of the information display platform, also the audience of the recommended information), and every click on the recommended information. Among them, the exposure in CPM refers to the display of the recommended information, and 1000 exposures means 1000 displays. The predetermined action in CPA is agreed upon by the service party and the information display platform in advance, for example, it can be registration, interaction, download, order, purchase, etc. When these composite indicators fluctuate greatly, it is necessary to locate the source of the fluctuation, and then make targeted subsequent adjustments.

[0034] According to the definitions of the above-mentioned composite indicators, the composite indicator is a ratio of the display resource and the interaction amount, wherein the interaction amount can be the thousand times of exposure (obtained by dividing the exposure by 1000, for example, the exposure is 10000, and the thousand times of exposure is 10, corresponding to CPM), the action amount (corresponding to CPA, indicating the number of times that the user performs a predetermined action), or the click amount (corresponding to CPC, indicating the number of times that the user clicks the recommended information). It can be seen that, although the composite indicator does not have the additivity, that is, for a batch of recommended information that needs to be analyzed (for example, in the period from 20 o'clock to 22 o'clock, all the recommended information displayed on the entire information display platform), the composite indicator of the recommended information cannot be obtained by adding the composite indicators of the influencing factors corresponding to the same dimension (for example, industry, business party, etc.), but the display resource and the interaction amount constituting the composite indicator belong to the additivity and can be added according to different dimensions, for example, can be added according to different industries and different business parties. The composite indicator fluctuation attribution scheme according to the example embodiments of the present disclosure utilizes the additivity of the display resource and the interaction amount, respectively calculates the first influence degree (reflecting the influence of the target factor on the fluctuation of the display resource) and the second influence degree (reflecting the influence of the target factor on the fluctuation of the interaction amount) of a plurality of target factors of the same dimension, and obtains the composite indicator influence degree of the corresponding target factor, so that the quantitative attribution of the composite indicator can be converted into the quantitative attribution of the additivity, the quantitative analysis of the composite indicator fluctuation is provided, and the source causing the composite indicator fluctuation is effectively determined from the plurality of target factors. The scheme can help the information display platform and the business party to understand the current situation, effectively explain the composite indicator, realize the quantitative attribution of the composite indicator fluctuation, improve the analysis efficiency, comprehensively consider the influences of the display resource and the interaction amount, and ensure the comprehensiveness and reliability of the analysis, so as to quickly locate the subdivision dimension causing the abnormal change of the composite indicator. In addition, the scheme can be applied to various scenes, and the composite indicator influence degree of different dimensions can be split according to different composite indicators and different dimensions, and has flexible and universal applicability.

[0035] In the following, the composite indicator fluctuation attribution method and the composite indicator fluctuation attribution device according to the example embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. Figures 1 to 5 The composite indicator fluctuation attribution method and the composite indicator fluctuation attribution device according to the example embodiments of the present disclosure will be described in detail with reference to the accompanying drawings.

[0036] Figure 1 FIG. 1 is a flow chart illustrating the composite indicator fluctuation attribution method according to the example embodiments of the present disclosure. Figure 2 FIG. 1 is a flow chart illustrating the composite indicator fluctuation attribution method according to the example embodiments of the present disclosure. Figure 3is another flowchart illustrating a composite index fluctuation attribution method according to an example embodiment of the present disclosure. It should be understood that the composite index fluctuation attribution method according to an example embodiment of the present disclosure can be implemented in a terminal device such as a smartphone, a tablet computer, a personal computer (PC), or in a device such as a server.

[0037] Referring to Figure 1 In step 101, a composite index corresponding to the recommendation information is determined, and target factors associated with the recommendation information are obtained, the composite index being used to measure a ratio between a display resource and an interaction amount corresponding to the recommendation information.

[0038] The recommendation information can be at least part of the recommendation information published on the information display platform, for example, all the recommendation information displayed by the information display platform in a time period of 20:00 to 22:00, for example, all the recommendation information displayed by the information display platform throughout the day, for example, all the recommendation information displayed by the information display platform throughout the day on a specific medium (such as a video). In other words, it is determined which composite index corresponding to which recommendation information needs to be analyzed, and target factors associated with these recommendation information are obtained.

[0039] The target factors are associated with the recommendation information, for example, can be associated with an object using the recommendation information. The object using the recommendation information can include a business party and an audience, the business party makes business recommendations through the recommendation information, and the audience can browse the recommendation information to obtain relevant information. The operation of obtaining the target factors associated with the recommendation information in step 101 includes: determining a target dimension associated with the recommendation information; and obtaining target factors under the target dimension based on the target dimension. The target dimension can be a dimension on the business party side or a dimension on the audience side. The business party invests display resources in the information display platform to publish the recommendation information, which will affect the display resources, and the audience is a user of the information display platform, as a potential user of the business corresponding to the recommendation information, which will affect the interaction amount. By doing fluctuation attribution from these two dimensions, it is helpful to fundamentally find the reason for the fluctuation of the composite index. By selecting the target dimension to be analyzed and further obtaining the target factors to be analyzed, targeted fluctuation attribution can be performed. It should be understood that for the acquisition of the target factors, at least part of the target factors under the same target dimension can be obtained to realize attribution analysis of the target dimension, or target factors of different target dimensions can be obtained to realize corresponding attribution analysis according to attribution requirements, which is not limited by the present disclosure.

[0040] As an example, the dimensions on the business side can include at least one of the following: industry, business. The industry refers to the industry to which each business on the information display platform belongs, and by performing attribution on the industry dimension, industry common problems can be found. The business refers to a specific subject, which helps to find individualized problems of a single business. It can be understood that each recommendation information corresponds to a unique business, and a business can publish at least one recommendation information. By specifically performing fluctuation attribution from these two dimensions, fluctuation reasons can be analyzed from different levels.

[0041] As an example, the dimensions on the audience side can include at least one of the following: user age, user gender, user region. Compared with the business, the number of audiences is large. By extracting representative features of the audience and taking them as attribution dimensions for fluctuation attribution, it is helpful to perform multi-angle macro analysis on the audience, find common problems of the audience, and improve analysis pertinence and analysis efficiency.

[0042] In step 102, the first influence degree and the second influence degree of the target factor are obtained. The first influence degree reflects the influence degree of the target factor on the fluctuation of the display resource of the recommendation information, and the second influence degree reflects the influence degree of the target factor on the fluctuation of the interaction amount of the recommendation information.

[0043] It should be understood that the influence degree here represents the fluctuation provided by the target factor. For all target factors of the same dimension, the sum of the first influence degrees of all target factors constitutes the fluctuation of the display resource, and the sum of the second influence degrees of all target factors constitutes the fluctuation of the interaction amount. For example, for the industry dimension, all target factors can be all industries involved in all recommendation information of the information display platform. The display resource and the interaction amount are additive indicators, so the influence of each target factor on the display resource change rate and the interaction amount change rate can be calculated, that is, the corresponding first influence degree and second influence degree, which are data preparation for subsequent calculation of the composite indicator influence degree. Its calculation is convenient, which guarantees the calculability of the composite indicator influence degree.

[0044] As an example, the first influence degree is the ratio of the growth amount of the target factor in the current period to the display resource relative to the base period display resource to the base period total display resource, and the base period total display resource is the sum of the base period display resources of all target factors of the same dimension. The second influence degree is the ratio of the growth amount of the target factor in the current period to the interaction amount relative to the base period interaction amount to the base period total interaction amount, and the base period total interaction amount is the sum of the base period interaction amounts of all target factors of the same dimension.

[0045] Optionally, the step of obtaining the second influence degree of the target factor in step 102 can include: obtaining any one of the thousand exposure amount influence degree, the action amount influence degree, and the click amount influence degree of the target factor. As can be known from the foregoing description, the thousand exposure amount, the action amount, and the click amount correspond to different composite indexes respectively, which means that the composite index fluctuation attribution method according to the example embodiments of the present disclosure can be applied to the fluctuation attribution of different composite indexes. When actually attributing, only the corresponding second influence degree and the data of the composite index need to be obtained according to the attribution requirement.

[0046] Taking the composite index CPM as an example, the CPM fluctuation is usually the week-on-week change of the large pool CPM obtained by summarizing all the recommended information successfully published on the information display platform. When the week-on-week change of the large pool CPM is large, it is hoped to be attributed to the industry, the business party, and other dimensions to help locate the source of the anomaly. Taking the industry attribution as an example, the first influence degree and the second influence degree of each specific industry (i.e., the target factor) in the large pool need to be calculated respectively.

[0047] For the first influence degree, as described above, the first influence degree = (the display resource in the current period - the display resource in the base period) / the total display resource in the base period, which can be expressed by the symbol as follows:

[0048]

[0049] wherein, Δcost i is the first influence degree of the industry i, cost t，i is the display resource of the industry i at the time t in the current period, cost t-7，i is the display resource of the industry i at the time t-7 in the base period, ∑ i cost t-7，i is the total display resource of all industries at the time t-7 in the base period.

[0050] In addition, the week-on-week change of the large pool display resource satisfies:

[0051]

[0052] Comparing formula (1) and formula (2), it can be found that the sum of the first influence degrees of all industries, that is, all target factors under the same target dimension, is equal to the change rate of the large pool display resource. Here, the change rate is specifically the week-on-week change. In practice, according to the statistical period of the composite index, the change rate can also be the day-on-day change, the month-on-month change, the year-on-year change, etc., which is not limited by the present disclosure.

[0053] For the interaction amount, the corresponding interaction amount of CPM is the thousand exposure amount, similar to the first influence degree, the second influence degree (specifically the thousand exposure amount influence degree) = (this period thousand exposure amount - base period thousand exposure amount) / base period total thousand exposure amount, at the same time, exposure amount influence degree = (this period exposure amount - base period exposure amount) / base period total exposure amount, the value of thousand exposure amount influence degree and exposure amount influence degree is equal, the second influence degree can be expressed by symbol as:

[0054]

[0055] Where, Δshow i is the second influence degree of industry i, show t，i is the exposure amount of industry i at this period t, show t-7，i is the exposure amount of industry i at the base period t-7, ∑ i show t-7，i is the total exposure amount of all industries at the base period t-7.

[0056] In addition, the large board exposure amount week-on-week satisfies:

[0057]

[0058] Like the display resource, the total of the second influence degrees of all industries, that is, all target factors under the same target dimension, is equal to the change rate of the large board exposure amount.

[0059] It can be understood that the first influence degree and the second influence degree of each target factor can be calculated in advance by other devices, and the execution subject (such as the aforementioned smart phone, server, etc.) of the method directly acquires in step 102, or the execution subject of the method calculates, and the present disclosure does not limit this.

[0060] At step 103, a composite index influence degree of the target factor is determined according to the first influence degree and the second influence degree. The composite index influence degree reflects the influence of the target factor on the fluctuation of the composite index of the recommended information. Similar to the first influence degree and the second influence degree, the sum of the composite index influence degrees of all target factors in the same dimension constitutes the fluctuation of the composite index. As described above, the display resource and the interaction amount belong to the summable index, and thus the first influence degree and the second influence degree of each target factor can be directly calculated by using the display resource data and the interaction amount data according to the formula (1) and the formula (3). The composite index is the ratio of the display resource and the interaction amount, and does not have the summable performance, and thus cannot be directly calculated by using the composite index data according to the calculation manner similar to the formula (1) and the formula (3). By changing the quantitative attribution of the composite index into the quantitative attribution of the two summable indexes of the display resource and the interaction amount, the first influence degree and the second influence degree are used to determine the composite index influence degree, which can provide the possibility for the quantitative analysis of the fluctuation of the composite index, and help to effectively determine the source of the fluctuation of the composite index from the multiple target factors. The scheme can realize the quantification of the attribution of the fluctuation of the composite index, help to improve the analysis efficiency, comprehensively consider the influences of the display resource and the interaction amount, and thus ensure the comprehensiveness and reliability of the analysis, and quickly locate the target factor causing the abnormal change of the composite index.

[0061] Optionally, referring to Figure 2 and Figure 3 , step 103 can be executed by determining the difference between the first influence degree and the second influence degree, and determining the composite index influence degree based on the difference.

[0062] Specifically, similar to the first influence degree and the second influence degree, the composite index influence degree is the ratio of the growth amount of the current period composite index relative to the base period composite index to the sum of the base period composite indexes of all target factors in the same dimension. Still taking the composite index CPM as an example, the CPM influence degree can be expressed as:

[0063]

[0064] Combined with the formula (1), (3), (5) and the definition of CPM, that is, CPM=cost / showx1000, it can be found through parameter relationship analysis that:

[0065] ΔCPM i =Δcost i -Δshow i -Δcost i ×Δshow i Formula (6)

[0066] wherein, Δcosti x Δshow i is the product of two percentages, whose absolute value is smaller, thus can be taken as the error term. Ignoring the error term, we can get:

[0067] ΔCPM i = Δcost i - Δshow i Equation (7)

[0068] The same applies to other composite indicators, so we can conclude that the composite indicator impact degree = first impact degree - second impact degree - first impact degree x second impact degree, and we can get the simplified relationship that approximately holds: composite indicator impact degree = first impact degree - second impact degree.

[0069] Based on the above relationship, the composite indicator impact degree of each target factor can be calculated according to the first impact degree and the second impact degree of each target factor according to different calculation requirements.

[0070] Referring to Figure 2 As an example, referring to Equation (7), the step of determining the composite indicator impact degree based on the difference value can specifically include: taking the difference value as the first composite indicator impact degree. By referring to the formula (7) after ignoring the error term, the calculation of the composite indicator impact degree can be simplified under the condition of ensuring the reliability of the result, which helps to reduce the operation load and improve the calculation efficiency.

[0071] Referring to Figure 3 As another example, referring to Equation (6), the step of determining the composite indicator impact degree based on the difference value can specifically include: obtaining the error term of the composite indicator impact degree; the error term reflects the error of the difference value relative to the composite indicator impact degree; based on the error term, the difference value is adjusted to obtain the second composite indicator impact degree. Specifically, the value obtained by subtracting the error term from the difference value is taken as the second composite indicator impact degree. It should be understood that the first composite indicator impact degree and the second composite indicator impact degree are distinguished here in order to reflect the difference in their calculation methods. In actual application, only one of them can be used, or both of them can be provided, and the specific one to be used can be selected by the staff. The present disclosure does not limit this. By referring to the accurate formula (6), the accuracy of the calculation result can be fully guaranteed to meet the fluctuation attribution requirements in high-precision scenarios. Specifically, still referring to Figure 4 Referring to Equation (6), the step of obtaining the error term of the composite indicator impact degree in this example can specifically be to determine the product of the first impact degree and the second impact degree as the error term of the composite indicator impact degree, which can clearly calculate the composite indicator impact degree and guarantee the calculation accuracy.

[0072] According to the composite index fluctuation attribution method of the example embodiment of the present disclosure, after the composite index influence degree is determined, further comprising: selecting at least one attribution factor from the plurality of target factors according to the composite index influence degree, wherein the composite index influence degree of the attribution factor is greater than the composite index influence degree of other target factors, or the absolute value of the composite index influence degree of the attribution factor is greater than the absolute value of the composite index influence degree of other target factors. By selecting the attribution factor with relatively large composite index influence degree or absolute value, the quantitative attribution diagnosis of the composite index fluctuation can be made, and it is considered that the change of the attribution factor is the main cause of the composite index fluctuation, which helps to provide a reference for further analysis of the composite index fluctuation.

[0073] It can be understood that, according to the analysis of the display resource and the interaction amount, similarly, the sum of the composite index influence degrees of all target factors under the same target dimension is equal to the change rate of the composite index, specifically the change rate of the current period composite index relative to the base period composite index. And according to formulas (6) and (7), the composite index influence degree of each target factor can be positive or negative, and the positive or negative can reflect the positive or negative influence of the corresponding target factor on the change rate of the composite index, and the absolute value can reflect the relative size of the composite index influence degree, so as to quickly lock the problem target factor to be detected and improve the efficiency of display resource attribution investigation. Specifically, when the change rate is negative, that is, the current period composite index decreases, the target factor with the minimum or smaller value of the composite index influence degree (i.e. negative value, and the absolute value is the largest or larger among all negative composite index influence degrees) is considered as the attribution factor, that is, the fluctuation source. When the change rate is positive, that is, the current period composite index increases, the target factor with the maximum or larger value of the composite index influence degree (i.e. positive value, and the absolute value is the largest or larger among all positive composite index influence degrees) is considered as the attribution factor.

[0074] As an example, to display the attribution factor, the composite index influence degree of each target factor can be directly outputted for subsequent analysis by the staff according to the output data. Of course, the first influence degree, the second influence degree, etc. can also be outputted together. For the embodiment using the first composite index influence degree, the error term (its calculation method is referred to formula (6) and the foregoing related embodiments) can be additionally calculated and outputted for the staff to fully understand the attribution situation.

[0075] As an example, to display the attribution factor, each target factor can also be arranged in ascending order or descending order according to the composite index influence degree, so as to intuitively locate the attribution factor with the largest influence or the attribution factors with larger influence, which is considered as the source of the composite index fluctuation. The composite index influence degree can also be outputted together when arranging the target factors, and the first influence degree, the second influence degree, the error term (also for the embodiment using the first composite index influence degree) etc. can be further outputted for the staff to fully understand the attribution situation.

[0076] The present disclosure does not limit the specific way of showing attribution factors, as long as the attribution situation can be known by the staff.

[0077] In one specific embodiment, assuming that the CPM on the tth day is -5.39% compared with the CPM on the t-7th day, the first and second influence degrees (specifically, the exposure influence degree) of each industry need to be calculated first, and then the CPM influence degree is determined according to the two influence degrees, and the calculation results are summarized as follows in Table 1.

[0078] Table 1 Summary of CPM Influence Degrees of Different Industries

[0079] Industry First Impact Degree Second Impact Degree CPM Impact Degree Error Term Industry 1 -4.54％ -1.98％ -2.56％ 0.09％ Industry 2 2.02％ 3.98％ -1.96％ 0.08％ Industry 3 -0.92％ 0.08％ -1.00％ 0.00％ Industry 4 0.67％ -0.06％ 0.73％ 0.00％ Industry 5 -1.18％ -0.49％ -0.69％ 0.01％ Industry 6 1.81％ 1.23％ 0.58％ 0.02％ Industry 7 -0.82％ -0.45％ -0.37％ 0.00％ Industry 8 0.51％ 0.82％ -0.31％ 0.00％ Industry 9 0.10％ 0.39％ -0.29％ 0.00％ Industry 10 0.73％ 0.44％ 0.29％ 0.00％ Industry 11 -0.37％ -0.24％ -0.13％ 0.00％ Industry 12 0.28％ 0.19％ 0.09％ 0.00％ Industry 13 0.31％ 0.28％ 0.03％ 0.00％ Industry 14 -0.03％ -0.01％ -0.02％ 0.00％ Industry 15 0.02％ 0.00％ 0.02％ 0.00％ Industry 16 -0.01％ -0.02％ 0.01％ 0.00％ Industry 17 -0.01％ -0.01％ 0.00％ 0.00％ Industry 18 0.00％ 0.01％ -0.01％ 0.00％ Industry 19 -0.01％ 0.00％ -0.01％ 0.00％ Industry 20 0.00％ 0.00％ 0.00％ 0.00％ Industry 21 0.04％ 0.04％ 0.00％ 0.00％ Industry 22 0.02％ 0.02％ 0.00％ 0.00％ Summation -0.38％ 4.22％ -5.60％ 0.21％

[0080] As can be seen from Table 1, for the week-on-week of the composite index CPM, the first influence degree of industry 1 is -4.54%, the exposure influence degree is -1.98%, and therefore the first CPM influence degree is:

[0081] The first CPM influence degree of industry 1 = -4.54% - (-1.98%) = -2.56%.

[0082] The error term of the CPM influence degree of industry 1 = |(-4.54%) * (-1.98%) | = 0.09%

[0083] The CPM influence degrees of each industry add up to -5.60%, which is approximately equal to the CPM week-on-week -5.39%. It can be considered that the decline of the CPM on the tth day is mainly caused by the decline of the CPM of industry 1, thereby quickly and efficiently locking the industry that needs to be investigated.

[0084] Figure 4 is a block diagram showing a composite index fluctuation attribution apparatus according to an example embodiment of the present disclosure. It should be understood that the composite index fluctuation attribution apparatus according to the example embodiment of the present disclosure can be implemented in a terminal device such as a smartphone, a tablet, a personal computer (PC) in software, hardware, or a combination of software and hardware, or can be implemented in a device such as a server.

[0085] Referring to Figure 4 , the composite index fluctuation attribution apparatus 200 includes a first acquisition unit 201, a second acquisition unit 202, and a determination unit 203.

[0086] The first acquisition unit 201 can determine a composite index corresponding to recommendation information, and acquire a target factor associated with the recommendation information, the composite index being used to measure a ratio between a display resource and an interaction amount corresponding to the recommendation information.

[0087] The recommended information can be at least part of the recommended information published on the information display platform, for example, all recommended information displayed by the information display platform in the period of 20 to 22 o'clock, or all recommended information displayed by the information display platform throughout the day, or all recommended information displayed by the information display platform on a specific medium (such as video) throughout the day. In other words, determine which composite indicators correspond to which recommended information, and obtain the target factors associated with these recommended information.

[0088] The target factor is associated with the recommended information, for example, it can be associated with the object using the recommended information. The object using the recommended information can include business parties and audiences. The business party makes business recommendations through the recommended information, and the audience can browse the recommended information to obtain related information. The operation of the first acquisition unit 201 to obtain the target factor associated with the recommended information includes: determining the target dimension associated with the recommended information; and based on the target dimension, obtaining the target factor under the target dimension. The target dimension can be a dimension on the business side, or a dimension on the audience side. The business party invests display resources to publish recommended information on the information display platform, which will affect the display resources. The audience is a user of the information display platform, as a potential user of the recommended information corresponding business, which will affect the interaction amount. By doing fluctuation attribution from these two dimensions, it is helpful to fundamentally find out the reason for the fluctuation of the composite indicator. By selecting the target dimension to be analyzed, and further obtaining the target factor to be analyzed, targeted fluctuation attribution can be performed. It should be understood that for the acquisition of the target factor, at least part of the target factors under the same target dimension can be acquired to realize attribution analysis of the target dimension, or target factors of different target dimensions can be acquired to realize corresponding attribution analysis according to the attribution demand, and the present disclosure is not limited.

[0089] The second acquisition unit 202 can acquire the first influence degree and the second influence degree of the target factor. The first influence degree reflects the influence degree of the target factor on the fluctuation of the display resources of the recommended information, and the second influence degree reflects the influence degree of the target factor on the fluctuation of the interaction amount of the recommended information.

[0090] It should be understood that the influence degree here represents the fluctuation provided by the target factor. For all target factors of the same dimension, the sum of the first influence degrees of all target factors constitutes the fluctuation of the display resources, and the sum of the second influence degrees of all target factors constitutes the fluctuation of the interaction amount. For example, for the industry dimension, all target factors can be all industries involved in all recommended information of the information display platform. The display resources and the interaction amount belong to additive indicators, so the influence of each target factor on the display resource change rate and the interaction amount change rate can be calculated, that is, the corresponding first influence degree and the second influence degree, as data preparation for subsequent calculation of the influence degree of the composite indicator. Its calculation is convenient, and ensures the calculability of the influence degree of the composite indicator.

[0091] As an example, the first influence degree is a ratio of a growth amount of the current period display resource relative to a base period display resource to a base period total display resource, the base period total display resource being a sum of base period display resources of all target factors of the same dimension, and the second influence degree is a ratio of a growth amount of the current period interaction amount relative to a base period interaction amount to a base period total interaction amount, the base period total interaction amount being a sum of base period interaction amounts of all target factors of the same dimension.

[0092] Optionally, the second acquisition unit 202 acquiring the second influence degree of the target factor can include: acquiring any one of a thousand exposure amount influence degree, an action amount influence degree, and a click amount influence degree of the target factor. As known from the foregoing, the thousand exposure amount, the action amount, and the click amount correspond to different composite indexes respectively, which means that the composite index fluctuation attribution method according to the example embodiments of the present disclosure can be applied to the fluctuation attribution of different composite indexes, and when actually attributing, only the corresponding second influence degree and the data of the composite index need to be acquired according to the attribution requirement.

[0093] It can be understood that the first influence degree and the second influence degree of each target factor can be calculated in advance by other devices, and the execution subject (such as the aforementioned smart phone, server, etc.) of the present method can directly acquire the first influence degree and the second influence degree in step 102, or the execution subject of the present method can calculate the first influence degree and the second influence degree, and the present disclosure does not limit this.

[0094] The determination unit 203 can determine a composite index influence degree corresponding to the target factor according to the first influence degree and the second influence degree. The composite index influence degree reflects the influence of the target factor on the fluctuation of the composite index of the recommended information. Similar to the first influence degree and the second influence degree, the sum of the composite index influence degrees of all target factors of the same dimension constitutes the fluctuation of the composite index. As described above, the display resource and the interaction amount belong to additive indexes, and thus the first influence degree and the second influence degree of each target factor can be directly calculated using the display resource data and the interaction amount data. The composite index is a ratio of the display resource and the interaction amount, and does not have the additive property, and thus the composite index influence degree cannot be directly calculated using the composite index data similar to the display resource and the interaction amount. By converting the quantitative attribution of the composite index as the composite index into the quantitative attribution of the display resource and the interaction amount as the two additive indexes, the first influence degree and the second influence degree are used to determine the composite index influence degree, which can provide the possibility for the quantitative analysis of the composite index fluctuation, and help to effectively determine the source causing the composite index fluctuation from the multiple target factors. This scheme can not only realize the quantification of the attribution of the composite index fluctuation, but also help to improve the analysis efficiency, and comprehensively consider the influence of the display resource and the interaction amount, thereby ensuring the comprehensiveness and reliability of the analysis, and quickly positioning the target factor causing the abnormal change of the composite index.

[0095] Optionally, the determining unit 203 can determine a difference between the first influence degree and the second influence degree, and determine the composite index influence degree based on the difference.

[0096] Specifically, similar to the first influence degree and the second influence degree, the composite index influence degree is a ratio of an increase amount of the current composite index relative to a base period composite index to a sum of base period composite indexes of all target factors of the same dimension. Through parameter relationship analysis, it can be found that the composite index influence degree = the first influence degree - the second influence degree - the first influence degree x the second influence degree, and an approximate simplified relationship formula can be obtained: the composite index influence degree = the first influence degree - the second influence degree.

[0097] Based on the above relationship formula, the composite index influence degree of each target factor can be calculated according to the first influence degree and the second influence degree of each target factor according to different calculation requirements.

[0098] As an example, the operation of determining the composite index influence degree based on the difference by the determining unit 203 can specifically include: taking the difference as a first composite index influence degree. By referring to the simplified relationship formula after ignoring the error term, the calculation of the composite index influence degree can be simplified under the condition of ensuring that the result is basically reliable, which helps to reduce the operation load and improve the calculation efficiency.

[0099] As another example, the operation of determining the composite index influence degree based on the difference by the determining unit 203 can specifically include: obtaining an error term of the composite index influence degree; the error term reflects an error of the difference relative to the composite index influence degree; and adjusting the difference based on the error term to obtain a second composite index influence degree. Specifically, a value obtained by subtracting the error term from the difference is taken as the second composite index influence degree. It should be understood that the first composite index influence degree and the second composite index influence degree are distinguished here to reflect that the calculation methods of the two are different. In actual application, only one of them can be used, or both of them can be provided, and the specific one to be used can be selected by the staff, and the present disclosure does not limit this. By referring to the accurate relationship formula, the accuracy of the calculation result can be fully guaranteed, and the fluctuation attribution requirement in the high-precision requirement scenario can be met. Specifically, the operation of obtaining the error term of the composite index influence degree by the determining unit 203 in this example can be specifically to determine the product of the first influence degree and the second influence degree as the error term of the composite index influence degree, which can clearly show how to calculate the composite index influence degree and guarantee the calculation accuracy.

[0100] The composite index fluctuation attribution device according to the example embodiment of the present disclosure can further comprise an attribution unit configured to select at least one attribution factor from the plurality of target factors according to the composite index influence degree, wherein the attribution factor has a composite index influence degree greater than the composite index influence degrees of other target factors, or has a composite index influence degree absolute value greater than the composite index influence degree absolute values of other target factors. By selecting the attribution factor with a relatively large composite index influence degree or composite index influence degree absolute value, quantitative attribution diagnosis can be performed on the composite index fluctuation, and it is considered that the fluctuation of the composite index is mainly caused by the change of the attribution factor, which helps to provide a reference for further analysis of the composite index fluctuation.

[0101] It can be understood that, according to the analysis of the display resource and the interaction amount, similarly, the sum of the composite index influence degrees of all target factors under the same target dimension is equal to the change rate of the composite index, specifically, the change rate of the current period composite index relative to the base period composite index. And according to the calculation relationship of the composite index influence degree, the composite index influence degree of each target factor can be positive or negative, and the positive and negative can reflect the positive and negative influence of the corresponding target factor on the change rate of the composite index, and the absolute value can reflect the relative size of the composite index influence degree, so as to quickly lock the problem target factor to be detected and improve the efficiency of display resource attribution investigation. Specifically, when the change rate is negative, that is, the current period composite index decreases, it can be considered that the target factor with the minimum or smaller value of the composite index influence degree (i.e. negative value, and the absolute value is the largest or larger among all negative composite index influence degrees) is the attribution factor, that is, the fluctuation source. When the change rate is positive, that is, the current period composite index increases, it can be considered that the target factor with the maximum or larger value of the composite index influence degree (i.e. positive value, and the absolute value is the largest or larger among all positive composite index influence degrees) is the attribution factor.

[0102] As an example, the attribution unit outputs the attribution factor, which can directly output the composite index influence degree of each target factor, so that the staff can perform subsequent analysis according to the output data. Of course, the first influence degree, the second influence degree, etc. can also be output together, and for the embodiment using the first composite index influence degree, an error term (its calculation method is referred to the related embodiment described above) can be additionally calculated and output for the staff to fully understand the attribution situation.

[0103] As an example, the attribution unit outputs the attribution factor, which can also arrange each target factor in ascending order or descending order according to the composite index influence degree, so as to intuitively locate the attribution factor with the largest influence or the multiple attribution factors with larger influence, which is considered to be the source of the composite index fluctuation. The composite index influence degree can also be output together when arranging the target factors, and the first influence degree, the second influence degree, the error term (also for the embodiment using the first composite index influence degree) etc. can be further output for the staff to fully understand the attribution situation.

[0104] The present disclosure does not limit the specific way in which the attribution unit presents attribution factors, as long as the attribution situation can be understood by the staff.

[0105] Figure 5 is a block diagram of an electronic device according to an exemplary embodiment of the present disclosure.

[0106] Referring to Figure 5 The electronic device 300 includes at least one memory 301 having a set of computer executable instructions stored therein and at least one processor 302, which, when executing the set of computer executable instructions, performs a composite indicator fluctuation attribution method according to an exemplary embodiment of the present disclosure.

[0107] As an example, the electronic device 300 can be a PC computer, a tablet device, a personal digital assistant, a smart phone, or other devices capable of executing the above-mentioned instruction set. Here, the electronic device 300 is not necessarily a single electronic device, but can be a collection of any devices or circuits capable of executing the above-mentioned instructions (or instruction sets) individually or jointly. The electronic device 300 can also be part of an integrated control system or a system manager, or can be configured as a portable electronic device that interfaces with a local or remote (e.g., via wireless transmission).

[0108] In the electronic device 300, the processor 302 can include a central processor (CPU), a graphics processor (GPU), a programmable logic device, a dedicated processor system, a microcontroller, or a microprocessor. As an example, but not limited to, the processor can also include an analog processor, a digital processor, a microprocessor, a multi-core processor, a processor array, a network processor, etc.

[0109] The processor 302 can run instructions or codes stored in the memory 301, wherein the memory 301 can also store data. Instructions and data can also be sent and received through a network via a network interface device, wherein the network interface device can use any known transmission protocol.

[0110] The memory 301 can be integrated with the processor 302, for example, arranging RAM or flash memory within an integrated circuit microprocessor, etc. In addition, the memory 301 can include a separate device, such as an external disk drive, a storage array, or other storage devices that can be used by any database system. The memory 301 and the processor 302 can be operatively coupled or can communicate with each other, for example, through I / O ports, network connections, etc., so that the processor 302 can read files stored in the memory.

[0111] In addition, the electronic device 300 can further include a video display such as a liquid crystal display and a user interaction interface such as a keyboard, a mouse, a touch input device, etc. All components of the electronic device 300 can be connected to each other via a bus and / or a network.

[0112] According to an exemplary embodiment of the present disclosure, there can also be provided a computer-readable storage medium that, when instructions in the computer-readable storage medium are executed by at least one processor, causes the at least one processor to perform a composite indicator fluctuation attribution method according to an exemplary embodiment of the present disclosure. Examples of the computer-readable storage medium here include read-only memory (ROM), random access programmable read-only memory (PROM), electrically erasable programmable read-only memory (EEPROM), random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), flash memory, non-volatile memory, CD-ROM, CD-R, CD+R, CD-RW, CD+RW, DVD-ROM, DVD-R, DVD+R, DVD-RW, DVD+RW, DVD-RAM, BD-ROM, BD-R, BD-R LTH, BD-RE, Blu-ray or optical disc memory, a hard disk drive (HDD), a solid state drive (SSD), a card-type memory such as a multimedia card, a secure digital (SD) card, or an extreme digital (XD) card, a magnetic tape, a floppy disk, a magneto-optical data storage device, an optical data storage device, a hard disk, a solid state disk, and any other device configured to store a computer program and any associated data, data files, and data structures in a non-transitory manner and provide the computer program and any associated data, data files, and data structures to a processor or computer so that the processor or computer can execute the computer program. The computer program in the computer-readable storage medium described above can be executed in an environment deployed in a computer device such as a client, a host, an agent device, a server, etc., and, in one example, the computer program and any associated data, data files, and data structures are distributed over a networked computer system so that the computer program and any associated data, data files, and data structures are stored, accessed, and executed in a distributed manner by one or more processors or computers.

[0113] According to an exemplary embodiment of the present disclosure, there can also be provided a computer program product including computer instructions that, when executed by at least one processor, cause the at least one processor to perform a composite indicator fluctuation attribution method according to an exemplary embodiment of the present disclosure.

[0114] According to the composite index fluctuation attribution method and device according to the example embodiments of the present disclosure, the additive property of the display resource and the interaction amount is utilized, the first influence degree and the second influence degree of the plurality of target factors are respectively obtained, and the composite index influence degree of the corresponding target factor is obtained, the quantitative attribution of the composite index can be converted into the quantitative attribution of the additive index, the quantitative analysis of the composite index fluctuation is possible, and the source causing the composite index fluctuation can be effectively determined from the plurality of target factors. The scheme can realize the quantification of the attribution of the composite index fluctuation, help to improve the analysis efficiency, comprehensively consider the influence of the display resource and the interaction amount, and thus guarantee the comprehensiveness and reliability of the analysis, and the target factor causing the abnormal change of the composite index can be quickly located. In addition, the scheme can be applied to various scenes, different dimension influence degrees can be split according to different indexes and different dimensions, and the scheme has flexibility and universality.

[0115] Other embodiments of the present disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the features disclosed herein. This application is intended to cover any variations, uses or adaptations of the present disclosure following, in general, the principles of the present disclosure and including such departures from the present disclosure that come within known

[0116] It should be understood that the present disclosure is not limited to the precise structures herein described and illustrated in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the claims that follow.

Claims

1. A composite indicator volatility attribution method, characterized in that, The method comprises the following steps: determining the composite index corresponding to the recommendation information, and obtaining target factors associated with the recommendation information, wherein the composite index is used to measure the ratio between the display resource and the interaction amount corresponding to the recommendation information; obtaining the first influence degree and the second influence degree of the target factors, wherein the first influence degree reflects the influence of the target factors on the fluctuation of the display resource of the recommendation information, and the second influence degree reflects the influence of the target factors on the fluctuation of the interaction amount of the recommendation information; determining the difference value obtained by subtracting the second influence degree from the first influence degree, and determining the composite index influence degree based on the difference value, wherein the composite index influence degree reflects the influence of the target factors on the fluctuation of the composite index of the recommendation information; selecting at least one attribution factor from the target factors according to the composite index influence degree, wherein the composite index influence degree of the attribution factor is greater than the composite index influence degree of other target factors, or the absolute value of the composite index influence degree of the attribution factor is greater than the absolute value of the composite index influence degree of other target factors.

2. The composite indicator volatility attribution method of claim 1, wherein, The step of determining the composite index influence degree based on the difference value comprises: taking the difference value as the first composite index influence degree.

3. The composite indicator volatility attribution method of claim 1, wherein, The step of determining the composite index influence degree based on the difference value comprises: obtaining an error term of the composite index influence degree; the error term reflects the error of the difference value relative to the composite index influence degree; adjusting the difference value based on the error term to obtain the second composite index influence degree.

4. The composite indicator volatility attribution method of claim 3, wherein, The step of obtaining the error term of the composite index influence degree comprises: determining the product of the first influence degree and the second influence degree as the error term.

5. The composite indicator volatility attribution method of any one of claims 1 to 4, wherein, The step of obtaining the target factors associated with the recommendation information comprises: determining the target dimension associated with the recommendation information; obtaining the target factors under the target dimension based on the target dimension.

6. A composite indicator volatility attribution apparatus, characterized by, The composite index fluctuation attribution device comprises: a first obtaining unit configured to determine the composite index corresponding to the recommendation information, and obtain target factors associated with the recommendation information, wherein the composite index is used to measure the ratio between the display resource and the interaction amount corresponding to the recommendation information; a second obtaining unit configured to obtain the first influence degree and the second influence degree of the target factors, wherein the first influence degree reflects the influence of the target factors on the fluctuation of the display resource of the recommendation information, and the second influence degree reflects the influence of the target factors on the fluctuation of the interaction amount of the recommendation information; a determination unit configured to determine the difference value obtained by subtracting the second influence degree from the first influence degree, and determine the composite index influence degree based on the difference value, wherein the composite index influence degree reflects the influence of the target factors on the fluctuation of the composite index of the recommendation information; An attribution unit is configured to select at least one attribution factor from the plurality of target factors according to the composite index influence degree, wherein the composite index influence degree of the attribution factor is greater than the composite index influence degree of other target factors, or the absolute value of the composite index influence degree of the attribution factor is greater than the absolute value of the composite index influence degree of other target factors.

7. The composite indicator volatility attribution apparatus of claim 6, wherein, The determination unit is further configured to: take the difference as a first composite index influence degree.

8. The composite indicator volatility attribution apparatus of claim 6, wherein, The determination unit is further configured to: obtain an error term of the composite index influence degree; the error term reflects an error of the difference relative to the composite index influence degree; adjust the difference based on the error term to obtain a second composite index influence degree.

9. The composite indicator volatility attribution apparatus of claim 8, wherein, The determination unit is further configured to: determine a product of the first influence degree and the second influence degree as the error term.

10. A composite indicator volatility attribution apparatus as claimed in any one of claims 6 to 9, wherein, The first obtaining unit is further configured to: determine a target dimension associated with the recommendation information; obtain a target factor under the target dimension based on the target dimension.

11. An electronic device, comprising: comprise: at least one processor; at least one memory storing computer-executable instructions, wherein the computer-executable instructions, when executed by the at least one processor, cause the at least one processor to perform the composite index fluctuation attribution method according to any one of claims 1 to 5.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by at least one processor, the at least one processor is caused to perform the composite index fluctuation attribution method according to any one of claims 1 to 5.

13. A computer program product comprising computer instructions, characterized in that, The computer instructions are executed by at least one processor to implement the composite index fluctuation attribution method according to any one of claims 1 to 5.

Images