Attribution index determination method and apparatus, device, and storage medium

Abstract

The application discloses an attribution index determination method and device, equipment and a storage medium, which can be applied to various scenes such as cloud technology, artificial intelligence, intelligent transportation and Internet of Vehicles, and the method comprises the steps of determining a target multimedia service index of a multimedia object in a preset period; obtaining at least two multimedia association indexes corresponding to the target multimedia service index; constructing an index logical tree based on the target multimedia service index and the at least two multimedia association indexes; determining a target node based on the association indexes between each parent node and the corresponding child node in the index logical tree; the association indexes represent the influence degree of the corresponding child node on each parent node; and determining the target multimedia association index represented by the target node as the attribution index of the target multimedia service index. The application realizes quick and accurate determination of the attribution index of the target multimedia service index.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, and particularly relates to a cause index determination method and device, equipment and a storage medium. BACKGROUND

[0002] In related technologies, there are more studies on short-term fluctuation detection of multimedia service indexes, but there are fewer studies on long-term trend decline of multimedia service indexes. At present, the cause analysis of the trend decline of multimedia service indexes is usually manual analysis, which is time-consuming and inefficient.

[0003] Therefore, it is necessary to provide a cause index determination method, device, equipment and storage medium to quickly and accurately determine the cause index of a target multimedia service index in a downward trend. SUMMARY

[0004] The present application provides a cause index determination method, device, equipment and storage medium, which can quickly and accurately determine the cause index of a target multimedia service index in a downward trend.

[0005] In one aspect, the present application provides a cause index determination method, which comprises:

[0006] determining a target multimedia service index of a multimedia object in a preset time period, the target multimedia service index being in a downward trend in the preset time period;

[0007] obtaining at least two multimedia associated indexes corresponding to the target multimedia service index;

[0008] constructing an index logical tree based on the target multimedia service index and the at least two multimedia associated indexes; a root node of the index logical tree represents the target multimedia service index, nodes other than the root node in the index logical tree represent the at least two multimedia associated indexes, and branches of the index logical tree represent the association relationship between different indexes;

[0009] determining a target node based on the association coefficient between each parent node and the corresponding child node in the index logical tree; the association coefficient represents the influence degree of the corresponding child node on the each parent node;

[0010] taking the multimedia associated index represented by the target node as a target multimedia associated index, and determining the target multimedia associated index as the cause index of the target multimedia service index.

[0011] Another aspect provides a cause index determination device, which comprises:

[0012] The target multimedia service index determination module is configured to determine a target multimedia service index of the multimedia object in a preset time period, the target multimedia service index being in a downward trend in the preset time period.

[0013] The multimedia association index acquisition module is configured to acquire at least two multimedia association indexes corresponding to the target multimedia service index.

[0014] The index logical tree construction module is configured to construct an index logical tree based on the target multimedia service index and the at least two multimedia association indexes, a root node of the index logical tree representing the target multimedia service index, nodes other than the root node in the index logical tree representing the at least two multimedia association indexes, and branches of the index logical tree representing association relationships between different indexes.

[0015] The target node determination module is configured to determine a target node based on an association coefficient between each parent node and a corresponding child node in the index logical tree, the association coefficient representing an influence degree of the corresponding child node on the each parent node.

[0016] The attribution index determination module is configured to determine a multimedia association index represented by the target node as a target multimedia association index, and determine the target multimedia association index as an attribution index of the target multimedia service index.

[0017] In another aspect, an attribution index determination device is provided, the device comprising a processor and a memory, the memory storing at least one instruction or at least one program, the at least one instruction or the at least one program being loaded and executed by the processor to implement the attribution index determination method as described above.

[0018] In another aspect, a computer storage medium is provided, the computer storage medium storing at least one instruction or at least one program, the at least one instruction or the at least one program being loaded and executed by a processor to implement the attribution index determination method as described above.

[0019] In another aspect, a computer program product or computer program is provided, the computer program product or computer program comprising computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to cause the computer device to perform to implement the attribution index determination method as described above.

[0020] The attribution index determination method, device, equipment and storage medium provided by the present application have the following technical effects:

[0021] After determining the target multimedia service index of the multimedia object in a preset time period, the application obtains a plurality of multimedia associated indexes corresponding to the target multimedia service index; thereby the association relationship between the target multimedia service index and the multimedia associated indexes, and the association relationship between the multimedia associated indexes can be used to construct an index logical tree; then the target nodes which have greater influence on the root node can be determined according to the association indexes between each parent node and the corresponding child node in the index logical tree; and the target multimedia associated index represented by the target node is determined as the attribution index of the target multimedia service index; thereby the attribution index of the target multimedia service index can be quickly and accurately determined. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0023] Figure 1 is a schematic diagram of an attribution index determination system provided by an embodiment of the present application;

[0024] Figure 2 is a flowchart of an attribution index determination method provided by an embodiment of the present application;

[0025] Figure 3 is a flowchart of a method for determining a target multimedia service index of a multimedia object in a preset time period provided by an embodiment of the present application;

[0026] Figure 4 is a flowchart of a method for determining a target node based on the association indexes between each parent node and the corresponding child node in the above index logical tree provided by an embodiment of the present application;

[0027] Figure 5 is a flowchart of a method for determining the current association indexes between the current node and the corresponding child node provided by an embodiment of the present application;

[0028] Figure 6 is another flowchart of a method for determining the current association indexes between the current node and the corresponding child node provided by an embodiment of the present application;

[0029] Figure 7 is a trend change curve of an exposure index in a current period provided by an embodiment of the present application;

[0030] Figure 8is a current period click volume index trend change curve provided by an embodiment of the present application;

[0031] Figure 9 is a current period consumption index trend change curve provided by an embodiment of the present application;

[0032] Figure 10 is a current period CPM index trend change curve provided by an embodiment of the present application;

[0033] Figure 11 is a current period fill rate index trend change curve provided by an embodiment of the present application;

[0034] Figure 12 is a current period target multimedia service index and multimedia association index association graph provided by an embodiment of the present application;

[0035] Figure 13 is a historical period exposure volume index trend change curve provided by an embodiment of the present application;

[0036] Figure 14 is a historical period click volume index trend change curve provided by an embodiment of the present application;

[0037] Figure 15 is a historical period consumption index trend change curve provided by an embodiment of the present application;

[0038] Figure 16 is a historical period CPM index trend change curve provided by an embodiment of the present application;

[0039] Figure 17 is a historical period fill rate index trend change curve provided by an embodiment of the present application;

[0040] Figure 18 is a structure schematic diagram of an attribution index determination device provided by an embodiment of the present application;

[0041] Figure 19 is a structure schematic diagram of a server provided by an embodiment of the present application. DETAILED DESCRIPTION

[0042] The professional glossary involved in the embodiment is explained as follows:

[0043] PLR: piecewise linear representation, piecewise linear representation;

[0044] cox_staurt: trend test analysis method;

[0045] CPM: Cost Per Mille, the revenue generated by 1000 exposures of an advertisement.

[0046] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0047] It should be noted that the terms "first", "second" and the like in the specification and claims of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or server including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to the process, method, product or device.

[0048] Please refer to Figure 1 , Figure 1 is a schematic diagram of an attribution index determination system provided by an embodiment of the present application, as shown in Figure 1 , the attribution index determination system can at least include a server 01 and a client 02.

[0049] Specifically, in the embodiments of the present application, the server 01 can include a standalone server, or a distributed server, or a server cluster composed of multiple servers, and can also be a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and basic cloud computing services such as big data and artificial intelligence platforms. The server 01 can include a network communication unit, a processor, a memory and the like. Specifically, the server 01 can be configured to construct an index logical tree based on a target multimedia service index and at least two multimedia association indexes; and take the multimedia association index represented by a target node in the index logical tree as a target multimedia association index, and determine the target multimedia association index as an attribution index of the target multimedia service index.

[0050] Specifically, in the embodiments of the present application, the client 02 can include an entity device such as a smart phone, a desktop computer, a tablet computer, a notebook computer, a digital assistant, a smart wearable device, a smart speaker, a vehicle terminal, a smart television, etc., can include a software running in the entity device, for example, a web page provided by a service provider to a user, and can be an application provided by the service provider to the user. Specifically, the client 02 can be used for online query of an attribution index of a target multimedia business index.

[0051] The following describes an attribution index determination method of the present application, Figure 2 is a flowchart of an attribution index determination method provided by the embodiments of the present application, and the present specification provides method operation steps as described in the embodiments or flowcharts, but more or fewer operation steps can be included based on conventional or non-creative labor. The order of steps listed in the embodiments is only one of the many execution orders of the steps, and does not represent the only execution order. In actual system or server product execution, the method order shown in the embodiments or the drawings can be executed in sequence or in parallel (for example, in a parallel processor or multi-thread processing environment). Specifically as shown in Figure 2 the method can include:

[0052] S201: determining a target multimedia business index of a multimedia object in a preset time period, wherein the target multimedia business index is in a downward trend in the preset time period.

[0053] Specifically, in the embodiments of the present application, the multimedia object can include but is not limited to a multimedia advertisement, a video, an audio, etc., and the preset time period can be selected according to actual needs, for example, one week, one month, one year, etc. The target multimedia business index can be a multimedia business index of the multimedia object, for example, consumption (advertising revenue), exposure, click volume, CPM, fill rate, click rate, etc. The target multimedia business index is in a downward trend in the preset time period, that is, the data in the target multimedia business index data set corresponding to the target multimedia business index in the preset time period shows a downward trend.

[0054] In the embodiments of the present application, as shown in Figure 3 the determination of the target multimedia business index of the multimedia object in the preset time period includes:

[0055] S20101: obtaining at least two multimedia business indexes of the multimedia object in the preset time period.

[0056] In the embodiment of the present application, at least two multimedia service indicators of the multimedia object in the preset time period can be obtained first. These multimedia service indicators can be in different trends, including an upward trend, a downward trend and a fluctuation trend. Then, the target multimedia service indicator in the downward trend is determined. Thus, the reason for the decrease of the target multimedia service indicator can be analyzed.

[0057] In the embodiment of the present application, after the at least two multimedia service indicators of the multimedia object in the preset time period are obtained, the method further includes:

[0058] The multimedia service indicator data set corresponding to each multimedia service indicator is obtained.

[0059] Specifically, in the embodiment of the present application, the multimedia service indicator data set can include data of the multimedia service indicator at different time points in the preset time period. The time points can be in days, weeks, months, etc. as granularity. Taking weeks as granularity, the sum of the multimedia service indicator data in a week is taken as a time point.

[0060] Specifically, in the embodiment of the present application, the trend of the indicator can be studied in weeks as granularity. The multimedia service indicator data set is composed of data points. For example, using 84 days from 2021.1.11 to 2021.4.4, the data is aggregated by weeks, and 12 data points of income are obtained. In the first week, the average income of the multimedia object is 500,000, and the data point 1 is (1, 50). In the second week, the average income is 700,000, and the data point 2 is (2, 70). In the third week, the average income is 40,000, and the data point 3 is (3, 40). In this way, the multimedia service indicator data set is obtained.

[0061] Abnormal data in each multimedia service indicator data set is determined. The abnormal data is the data in the target time period.

[0062] Specifically, in the embodiment of the present application, the target time period can be a promotion period for the multimedia object, such as a special date corresponding period of Double Eleven, Six One Eight, etc. The data value in the target time period is usually large, and the difference with the data in the daily period is large. Therefore, it is determined as abnormal data.

[0063] The abnormal data in each multimedia service indicator data set is deleted to obtain an updated multimedia service indicator data set.

[0064] Specifically, in the embodiment of the present application, the abnormal data in each multimedia service indicator data set can be deleted, so as to avoid the influence of the abnormal data on the trend analysis of the multimedia service indicator.

[0065] In the embodiments of the present application, by screening out abnormal data in the multimedia service index data set, the accuracy of the multimedia service index trend analysis can be improved, thereby improving the accuracy of the determined target multimedia service index.

[0066] S20103: determining an initial change trend of each multimedia service index in the preset time period based on a trend test method;

[0067] Specifically, in the embodiments of the present application, the data in the multimedia service index data set of each multimedia service index in the preset time period can be divided into multiple sequences, and the trend test method is used to determine the change trend of each multimedia service index in the preset time period. The principle of the trend test method is as follows:

[0068] Suppose that a multimedia service index data set includes data sequences x1, x2, x3, …, xn. The sequence is divided into two parts with a constant c as a boundary, and the two parts are paired into the form of (x1, xc+1), (x2, xc+2), … (xc, xn), where c=n / 2 when n is even, and c=(n+1) / 2 when n is odd, and the middle number xc+1 is discarded. Then, the latter number in each group is subtracted from the former number, and the positive and negative are recorded. s+ represents the number of positive numbers, that is, the latter number is greater than the former number, and s- represents the number of negative numbers. p(+) represents the probability of taking a positive number, and p(-) represents the probability of taking a negative number. In this way, we obtain the sign test method to test whether the sequence has a trend. Under the null hypothesis of no trend, the data set follows a binomial distribution, that is, p(+) = 0.5.

[0069] In the embodiments of the present application, the determination of the initial change trend of each multimedia service index in the preset time period based on the trend test method includes:

[0070] Based on the trend test method, the initial change trend of the data in each updated multimedia service index data set in the preset time period is determined.

[0071] S20105: determining the multimedia service index with a downward trend as a candidate multimedia service index;

[0072] S20107: determining the slope of the trend change curve of the candidate multimedia service index based on the least square method;

[0073] In the embodiments of the present application, after the candidate multimedia service index in the downward trend is determined by the trend test method, the trend change curve of the candidate multimedia service index is obtained according to the candidate multimedia service index data set corresponding to the candidate multimedia service index, and the slope of the trend change curve of the candidate multimedia service index is calculated by using the least square method.

[0074] S20109: determining the candidate multimedia service index with the slope less than the preset slope threshold as the target multimedia service index.

[0075] In the embodiments of the present application, the preset slope threshold can be set according to actual conditions, for example, it can be set to -0.3, and when the slope of the curve is less than -0.3, it is determined that the candidate multimedia service index corresponding to the curve is the target multimedia service index in a downward trend.

[0076] Specifically, in the embodiments of the present application, if the multimedia service index is in an upward trend, the value of the data of the multimedia service index in the later period is significantly greater than the value of the data in the earlier period; otherwise, if the multimedia service index is in a downward trend, the value of the data in the later period is significantly less than the value of the data in the earlier period. The positive and negative of the difference between the data in the two periods can be used to determine the overall trend of the data.

[0077] In the embodiments of the present application, for the oscillating data, the least square method is prone to false judgment due to interference, which brings certain trouble to the selection of the preset slope threshold; that is, the error of the target multimedia service index determined by the least square method alone is large. Before the least square method, the cox_staurt method is used to make an initial trend judgment on all multimedia service indexes, which can well solve the error caused by the oscillating data, and the use of the cox_staurt method in combination with the least square method can improve the accuracy of determining the target multimedia service index.

[0078] S203: obtaining at least two multimedia associated indexes corresponding to the target multimedia service index.

[0079] Specifically, in the embodiments of the present application, the multimedia associated indexes and the target multimedia service index have a direct or indirect association. The multimedia associated indexes can be determined according to the calculation formula of the target multimedia service index; for example, the calculation formula is: target multimedia service index M=X*Y, X=A*B, Y=C*D.

[0080] Wherein, the influencing factor of index C is index E; then indexes X, Y, A, B, C, D, and E are all multimedia associated indexes of target multimedia service index M.

[0081] S205: constructing an index logical tree based on the target multimedia service index and the at least two multimedia associated indexes; the root node of the index logical tree represents the target multimedia service index, the nodes other than the root node in the index logical tree represent the at least two multimedia associated indexes, and the branches of the index logical tree represent the association between different indexes.

[0082] In the embodiments of the present application, the target multimedia service index can be represented by a root node to construct an index logical tree; indexes having a direct correlation relationship are connected by branches. The index logical tree includes a plurality of parent nodes and corresponding child nodes. The topmost node in the logical tree is the root node, which has no child nodes. The bottommost node is a leaf node, which also has no child nodes.

[0083] In the embodiments of the present application, the target node can be determined based on the correlation coefficient between each parent node and the corresponding child node in the index logical tree. The correlation coefficient represents the influence degree of the corresponding child node on each parent node.

[0084] In the embodiments of the present application, the target node can be determined by calculating the correlation coefficient of the parent node and the child node in the index logical tree. The target node is a leaf node in the index logical tree. The target node is the node that has the greatest influence on the root node. That is, the multimedia correlation index corresponding to the target node is the main reason for the decline of the target multimedia service index. The target node can be one or more. In the embodiments of the present application, the plurality refers to at least two.

[0085] In the embodiments of the present application, as shown in Figure 4 The determination of the target node based on the correlation coefficient between each parent node and the corresponding child node in the index logical tree includes the following steps.

[0086] S2071: Taking the root node of the index logical tree as the current node;

[0087] In the embodiments of the present application, the index logical tree can be traversed to obtain the root node as the first parent node, and the root node is taken as the current node.

[0088] S2073: Determining the current correlation coefficient between the current node and the corresponding child node;

[0089] In the embodiments of the present application, the current node can correspond to one or more child nodes. When there are a plurality of child nodes, the current correlation coefficient between each child node and the current node is calculated.

[0090] In the embodiments of the present application, as shown in Figure 5 The determination of the current correlation coefficient between the current node and the corresponding child node includes the following steps.

[0091] S20731: Based on the target multimedia service index data set corresponding to the target multimedia service index, a trend change curve of the target multimedia service index is constructed;

[0092] S20733: constructing a trend change curve of the first multimedia correlation index based on the first multimedia correlation index data set corresponding to the corresponding sub-node;

[0093] In the embodiments of the present application, the corresponding sub-node represents the first multimedia correlation index, and the trend change curve of the first multimedia correlation index is constructed based on the first multimedia correlation index data set corresponding to the corresponding sub-node, which includes:

[0094] The trend change curve of the first multimedia correlation index is constructed based on the first multimedia correlation index data set corresponding to the first multimedia correlation index.

[0095] In the embodiments of the present application, the first multimedia correlation index data set and the target multimedia service index data set are data in the same preset time period, and both have the same time granularity.

[0096] S20735: determining a first mode distance between the trend change curve of the target multimedia service index and the trend change curve of the first multimedia correlation index;

[0097] In the embodiments of the present application, the first mode distance between the trend change curves can be calculated by the PLR algorithm.

[0098] In the embodiments of the present application, the calculation principle of the PLR algorithm is as follows:

[0099] For a time series, the data change trend includes three states: rising, falling and constant, and the state is represented as M={1,-1,0}. Assuming that there is a sequence of length A, it is divided into K segments. For each segment, a slope is calculated, and the slope is positive, indicating rising, the slope is negative, indicating falling, and the slope is 0, indicating constant. Then the sequence can be represented as [1,1,0,-1...] such a sequence, and adjacent identical patterns are merged to obtain a sequence of [1,0,-1...].

[0100] The algorithm steps are as follows:

[0101] 1. Merge adjacent identical patterns to obtain a sequence of [1,0,-1...];

[0102] 2. For two curves S1, S2: adjacent identical patterns are merged, so the obtained pattern sequence is arranged at intervals of 1, -1, and 0, and each pattern can span different lengths of time. After merging the patterns, the sequence S1 can have N patterns, and S2 can have M patterns. They need to be equalized.

[0103] After PLR, S1 and S2 are represented as:

[0104] S1 = {(m 11 , t 11 ), …, (m 1N , t 1N )}

[0105] S2 = {(m 21 , t 21 ), …, (m 2M , t 2M )}

[0106] Let S1i, S2j represent the i, jth pattern of S1, S2, i.e. S1i = (m 1i , t 1i ), t represents time, and no matter how cut, the final end is equal, i.e. t iN = t 2M . The specific formula is as follows:

[0107] S1 = {(1, t 11 ), (-1, t 12 ), (0, t 13 )}

[0108] S2 = {(-1, t 21 ), (1, t 22 )}

[0109] After the mode is digitized, they are deformed as:

[0110] S1 = {(1, t1), (-1, t2), (-1, t3), (0, t4)}

[0111] S2 = {(1, t1), (1, t2), (-1, t3), (-1, t4)}

[0112] That is, let them use the common split point to obtain the mode sequence with equal length.

[0113] According to the mode distance formula:

[0114] D = |m 1i - m 2i |, obviously, D ∈ (0, 1, 2), the closer to 0, the more similar the mode; the closer to 2, the more dissimilar the mode. Add all the mode distances, i.e. the mode distance of the time sequence:

[0115]

[0116] However, it should be noted that each pattern can span different lengths of time, and the longer a pattern lasts, the more information it contains about the entire sequence, so weighting is required, and the updated pattern distance is calculated as follows:

[0117]

[0118] where t wi = t i / t N , t i is the length of time spanned by the i-th pattern, and t N is the total length of time.

[0119] S20737: Take the first pattern distance as the correlation coefficient between the current node and the corresponding child node.

[0120] In the embodiments of the present application, as shown in Figure 6 , the current node represents a second multimedia correlation indicator, and the determination of the current correlation coefficient between the current node and the corresponding child node includes:

[0121] S207301: Based on the second multimedia correlation indicator dataset corresponding to the second multimedia correlation indicator, a trend change curve of the second multimedia correlation indicator is constructed.

[0122] S207303: Based on the third multimedia correlation indicator dataset corresponding to the corresponding child node, a trend change curve of the third multimedia correlation indicator is constructed.

[0123] In the embodiments of the present application, the corresponding child node represents a third multimedia correlation indicator, and the construction of the trend change curve of the third multimedia correlation indicator based on the third multimedia correlation indicator dataset corresponding to the corresponding child node includes:

[0124] Based on the third multimedia correlation indicator dataset corresponding to the third multimedia correlation indicator, a trend change curve of the third multimedia correlation indicator is constructed.

[0125] In the embodiments of the present application, the second multimedia correlation indicator dataset and the third multimedia correlation indicator dataset are data in the same preset time period, and both use the same time granularity.

[0126] S207305: Determine a second pattern distance between the trend change curve of the second multimedia correlation indicator and the trend change curve of the third multimedia correlation indicator.

[0127] In the embodiments of the present application, the second pattern distance between the trend change curves can be calculated by the PLR algorithm.

[0128] S207307: taking the second mode distance as the correlation coefficient between the current node and the corresponding child node.

[0129] Specifically, in the embodiment of the present application, as shown in the figure, Figures 7-11 Figures 7-11 are the trend change curves corresponding to the exposure (times), click volume (times), consumption (yuan), CPM (yuan), and filling rate indicators of the current period, respectively.

[0130] Among them, the consumption shows a downward trend, with a decline of-9.28%, which is mainly related to CPM, and the correlation coefficient is 0.68;

[0131] CPM shows a downward trend, with a decline of-42.54%, which is mainly related to CPC, and the correlation coefficient is 0.95;

[0132] CPC shows a downward trend, with a decline of-26.15%;

[0133] The click volume shows a downward trend, with a decline of-22.13%;

[0134] The filling rate shows a downward trend, with a decline of-8.13%.

[0135] In daily work, the same ring rate of income-related indicators (core indicators) is easy to cause attention, but the long-term trend of small and medium-sized traffic indicators is often easily ignored. The present application can make up for the deficiency of traffic indicator monitoring, cover the detection of multimedia business indicators in a long-term downward trend, and guide the product and operation to pay attention to related problems.

[0136] In the embodiment of the present application, the mode distance between the indicators can be calculated through the trend change curve corresponding to the indicators, and the correlation coefficient between the current node and the corresponding child node is determined by calculating the mode distance, so as to facilitate further determination of the child node which has greater influence on the current node.

[0137] S2075: taking the child node with the current correlation coefficient greater than the preset threshold as the current node again;

[0138] In the embodiment of the present application, if the current correlation coefficient is less than or equal to the preset threshold, the influence of the current node on the root node is small, and the current node is excluded. If the current correlation coefficient is greater than the preset threshold, it means that the influence of the current node on the root node is large, and it is necessary to continue to judge whether it has child nodes.

[0139] S2077: if the current node has corresponding child nodes, repeat the steps of determining the current correlation coefficient between the current node and the corresponding child node, and taking the child node with the current correlation coefficient greater than the preset threshold as the current node again; ​

[0140] In the embodiments of the present application, if the current node has corresponding child nodes, it is necessary to continue to determine the current correlation coefficient of the current node and the corresponding child nodes, and determine the child nodes with the current correlation coefficient greater than the preset threshold, until the current node is a leaf node.

[0141] S2079: If the current node has no corresponding child nodes, the current node is determined as the target node.

[0142] Specifically, in the embodiments of the present application, the current node has no corresponding child nodes means that the current node is a leaf node in the last layer of the index logical tree.

[0143] In a specific embodiment, for example, in the application scenario where the multimedia object is a multimedia advertisement, as shown in the following table, the table is a correlation diagram of the target multimedia service index and the multimedia correlation index, wherein the target multimedia service index is consumption, and the calculation formula of the consumption is as follows: Figure 12

[0144] Consumption = Exposure * CPM Elimination;

[0145] Exposure = Request Number * Filling Rate * Display Rate;

[0146] Among them, the indexes having a direct correlation with the CPM are pCTR bias, pCVR bias, super-revenue ratio, and the number of target users; when the multimedia object is multiple, the target user is the investor of the multimedia object, wherein the target user can be a user corresponding to one or more multimedia objects in the multiple multimedia objects, and the user is a user who invests more cost in the multimedia object. The more the number of target users is, the greater the influence on the CPM is.

[0147] In the embodiments of the present application, CTR (Click-Through-Rate) is a commonly used term in Internet advertising, which refers to the click-through rate (click rate) of network advertising (picture advertising / text advertising / keyword advertising / ranking advertising / video advertising, etc.), that is, the actual number of clicks of the advertisement (strictly speaking, the number of target pages reached) divided by the display amount (Show content) of the advertisement.

[0148] Among them, pCTR bias is the deviation of CTR, and pCTR bias is associated with CTR; the calculation formula is as follows:

[0149] pCTR bias = (pCTR / CTR)-1;

[0150] In the formula, pCTR is the estimated click rate, and CTR is the real click rate; changing CTR will cause pCTR bias to change. ​

[0151] In the embodiments of the present application, the CVR (Conversion Rate) is the conversion rate from the user clicking on the advertisement to becoming an effective activation or registration or even a paying user. The calculation formula is as follows:

[0152] CVR=(conversion amount / click amount)*100%;

[0153] Wherein, pCVR bias is the bias of CVR, and pCVR bias is associated with CVR. The calculation formula is as follows:

[0154] pCVR bias=(pCVR / CVR)-1;

[0155] In the formula, pCVR is the estimated conversion rate, and CVR is the real conversion rate. Changing CVR will cause pCVR bias to change.

[0156] In the embodiments of the present application, the calculation formula of the super-revenue ratio is as follows:

[0157] Super-revenue ratio=(consumption-GMV) / GMV;

[0158] Wherein, the consumption is the advertising income, and GMV (Gross Merchandise Volume) represents the fee that the advertiser should pay to the advertising system. The calculation formula is as follows:

[0159] GMV=target bid of the advertiser*actual conversion number of the advertisement.

[0160] In the embodiments of the present application, the exposure amount is affected by the request number, the filling rate, the display rate and other indicators, and the request number is affected by the user number (user visitor, UV).

[0161] In the embodiments of the present application, the global fill rate represents the number of requests with advertisements returned by the advertising position / the total number of requests, indicating the utilization rate of the advertising position. The higher the fill rate, the greater the utilization rate. The calculation formula of the global fill rate is as follows:

[0162] Global fill rate=requests with advertisements returned by the advertising position (i.e. returned amount) / total requests of the advertising position;

[0163] In the embodiments of the present application, for example, the client sends an advertisement acquisition request to the advertising platform, the advertising platform returns an advertisement in response to the request, which is counted as one filling. After returning the advertisement, the client receives the filling, and the advertisement will be displayed, and the user will see the advertisement, which is counted as one display. The calculation formula of the display rate is as follows:

[0164] Display rate=ad exposure amount / ad return amount.

[0165] In the embodiments of the present application, when the index logical tree is constructed, the root node can represent the consumption in a downward trend, the correlation coefficients between the consumption and the exposure and the CPM are calculated first, for example, the correlation coefficient between the consumption and the exposure is 0.2, and the correlation coefficient between the consumption and the CPM is 0.8, then it is determined that the preliminary cause of the consumption decrease is the CPM; then the correlation coefficients between the CPM and each index (such as pCTR bias, pCVR bias, super-revenue ratio, and the number of target users) directly associated with the CPM are calculated, if the correlation coefficient between the super-revenue ratio index and the CPM is greater than a preset threshold, since the index has no child node, the index is determined as the attribution index of the consumption. If the correlation coefficient between the pCVR bias and the CPM is greater than the preset threshold, the correlation coefficient between the pCVR bias and the corresponding child node CVR is calculated, and if the calculation result is greater than the preset threshold, the CVR is determined as the attribution index of the consumption. In some other embodiments, the attribution index of the consumption can include multiple.

[0166] In the embodiments of the present application, after the current correlation coefficient between the current node and the corresponding child node is calculated, the child node with the current correlation coefficient greater than a preset threshold can be re-taken as the current node; and when the updated current node has a child node, the current correlation coefficient between the parent node and the child node is repeatedly calculated until the current node is a leaf node, and the current node is taken as the target node; so that the attribution index affecting the decrease of the target multimedia service index can be quickly found according to the index logical tree.

[0167] S209: taking the multimedia correlation index represented by the target node as a target multimedia correlation index, and determining the target multimedia correlation index as the attribution index of the target multimedia service index.

[0168] In the embodiments of the present application, the attribution index of the target multimedia service index can be one or more.

[0169] In the embodiments of the present application, the target node is at least two, and the target multimedia correlation index represented by the target node is taken as a target multimedia correlation index, and the target multimedia correlation index is determined as the attribution index of the target multimedia service index, including:

[0170] taking the multimedia correlation index represented by each target node as a target multimedia correlation index;

[0171] obtaining a historical index data set of each target multimedia correlation index in a historical period;

[0172] obtaining a current index data set of each target multimedia correlation index in the preset period;

[0173] Target multimedia-related indicators whose historical indicator data and current indicator data show consistent trends are identified as multimedia-related indicators to be screened out.

[0174] The indicators other than the multimedia-related indicators to be screened out from the above at least two target multimedia-related indicators are determined as the attribution indicators of the above target multimedia business indicators.

[0175] In this embodiment, if there are at least two target nodes, there are also at least two corresponding target multimedia association indicators. Historical data can be used to screen these indicators, deleting those that show a downward trend within the same preset time period in the historical context, thereby determining the attribution indicator. For example, if a multimedia association indicator shows a downward trend in both the preset time period of the year or several years prior to this year and the preset time period of this year, it indicates a normal decline and cannot be used as an attribution indicator. Similarly, during the back-to-school season each year, the click-through rate of applications (Apps) targeting students tends to decline, which is a normal decline and cannot be further improved.

[0176] In this application embodiment, metrics such as impressions (times), clicks (times), cost (yuan), CPM (yuan), and fill rate are determined as attribution metrics, and metric data for these metrics in the same historical period can be obtained; such as Figures 13-17 As shown, Figures 13-17 Historical indicator data from December 2019 to February 2020, compared with... Figures 7-11 Comparing the current indicator data from December 2020 to February 2021, it can be seen that CPM was also on a downward trend during the same period in the historical period. In this case, CPM is not used as an attribution indicator, thus improving the accuracy of determining the attribution indicator.

[0177] As can be seen from the technical solutions provided by the embodiments of this application above, after determining the target multimedia business indicator of the multimedia object in a downward trend within a preset time period, the embodiments of this application obtain multiple multimedia related indicators corresponding to the target multimedia business indicator; thereby, an indicator logic tree can be constructed based on the correlation between the target multimedia business indicator and the multimedia related indicators, as well as the correlation between the multimedia related indicators; then, based on the correlation coefficient between each parent node and its corresponding child node in the indicator logic tree, the target node that has a greater impact on the root node is determined; and the target multimedia related indicator represented by the target node is determined as the attribution indicator of the target multimedia business indicator; thereby achieving the rapid and accurate determination of the attribution indicator of the target multimedia business indicator.

[0178] This application also provides an attribution index determination device, such as... Figure 18 As shown, the device includes:

[0179] The target multimedia service index determination module 1810 is configured to determine a target multimedia service index of a multimedia object in a preset time period, where the target multimedia service index is in a downward trend in the preset time period.

[0180] The multimedia association index acquisition module 1820 is configured to acquire at least two multimedia association indexes corresponding to the target multimedia service index.

[0181] The index logical tree construction module 1830 is configured to construct an index logical tree based on the target multimedia service index and the at least two multimedia association indexes, where a root node of the index logical tree represents the target multimedia service index, nodes other than the root node in the index logical tree represent the at least two multimedia association indexes, and branches of the index logical tree represent association relationships between different indexes.

[0182] The target node determination module 1840 is configured to determine a target node based on an association coefficient between each parent node and a corresponding child node in the index logical tree, where the association coefficient represents an influence degree of the corresponding child node on the each parent node.

[0183] The attribution index determination module 1850 is configured to determine a multimedia association index represented by the target node as a target multimedia association index, and determine the target multimedia association index as an attribution index of the target multimedia service index.

[0184] In some embodiments, the target node determination module can include:

[0185] The current node determination unit is configured to determine a root node of the index logical tree as a current node.

[0186] The current association coefficient determination unit is configured to determine a current association coefficient between the current node and a corresponding child node.

[0187] The current node updating unit is configured to determine a child node with a current association coefficient greater than a preset threshold as a new current node.

[0188] The step repeating unit is configured to, if the current node has a corresponding child node, repeat the steps of determining a current association coefficient between the current node and a corresponding child node, and determining a child node with a current association coefficient greater than a preset threshold as a new current node.

[0189] The target node determination unit is configured to, if the current node has no corresponding child node, determine the current node as the target node.

[0190] In some embodiments, the current node represents the target multimedia service indicator, and the current correlation coefficient determination unit can include:

[0191] a trend change curve construction subunit for constructing a trend change curve of the target multimedia service indicator based on a target multimedia service indicator data set corresponding to the target multimedia service indicator;

[0192] a first curve construction subunit for constructing a trend change curve of the first multimedia correlation indicator based on a first multimedia correlation indicator data set corresponding to the corresponding child node;

[0193] a first mode distance determination subunit for determining a first mode distance between the trend change curve of the target multimedia service indicator and the trend change curve of the first multimedia correlation indicator;

[0194] a first correlation coefficient determination subunit for taking the first mode distance as a correlation coefficient between the current node and the corresponding child node.

[0195] In some embodiments, the current node represents a second multimedia correlation indicator, and the current correlation coefficient determination unit can include:

[0196] a second curve construction subunit for constructing a trend change curve of the second multimedia correlation indicator based on a second multimedia correlation indicator data set corresponding to the second multimedia correlation indicator;

[0197] a third curve construction subunit for constructing a trend change curve of a third multimedia correlation indicator based on a third multimedia correlation indicator data set corresponding to the corresponding child node;

[0198] a second mode distance determination subunit for determining a second mode distance between the trend change curve of the second multimedia correlation indicator and the trend change curve of the third multimedia correlation indicator;

[0199] a second correlation coefficient determination subunit for taking the second mode distance as a correlation coefficient between the current node and the corresponding child node.

[0200] In some embodiments, the target nodes are at least two, and the apparatus can further include:

[0201] In some embodiments, the attribution indicator determination module can include:

[0202] a target multimedia correlation indicator determination unit for taking the multimedia correlation indicator represented by each target node as a target multimedia correlation indicator;

[0203] a historical index data set acquisition unit, configured to acquire a historical index data set of each target multimedia association index in a historical period;

[0204] a current index data set unit, configured to acquire a current index data set of each target multimedia association index in the preset period;

[0205] a multimedia association index to be screened out determination unit, configured to determine, as a multimedia association index to be screened out, a target multimedia association index whose data in the historical index data set is consistent with a change trend of data in the current index data set;

[0206] an attribution index determination unit, configured to determine, as an attribution index of the target multimedia service index, an index other than the multimedia association index to be screened out in the at least two target multimedia association indexes.

[0207] In some embodiments, the target multimedia service index determination module can include:

[0208] a multimedia service index acquisition unit, configured to acquire at least two multimedia service indexes of the multimedia object in the preset period;

[0209] an initial change trend determination unit, configured to determine, based on a trend test method, an initial change trend of each multimedia service index in the preset period;

[0210] a candidate multimedia service index determination unit, configured to determine, as a candidate multimedia service index, a multimedia service index whose initial change trend is a downward trend;

[0211] a slope determination unit, configured to determine, based on a least square method, a slope of a trend change curve of the candidate multimedia service index;

[0212] a target multimedia service index determination unit, configured to determine, as the target multimedia service index, a candidate multimedia service index whose slope is less than a preset slope threshold.

[0213] In some embodiments, the apparatus can further include:

[0214] a multimedia service index data set acquisition module, configured to acquire a multimedia service index data set corresponding to each multimedia service index;

[0215] an abnormal data determination module, configured to determine abnormal data in each multimedia service index data set, the abnormal data being data in a target period;

[0216] an abnormal data deletion module, configured to delete the abnormal data in each multimedia service index data set, to obtain an updated multimedia service index data set.

[0217] In some embodiments, the initial change trend determination unit can comprise:

[0218] an initial change trend determination sub-unit configured to determine, based on the trend test method, an initial change trend of data in each updated multimedia service index data set within the preset time period.

[0219] The device in the device embodiment and the method embodiment are based on the same inventive concept.

[0220] The embodiments of the present application provide an attribution index determination device, which comprises a processor and a memory. The memory stores at least one instruction or at least one program, which is loaded and executed by the processor to implement the attribution index determination method provided in the above method embodiments.

[0221] The embodiments of the present application also provide a computer storage medium, which can be arranged in a terminal to save at least one instruction or at least one program related to the attribution index determination method in the method embodiments. The at least one instruction or at least one program is loaded and executed by the processor to implement the attribution index determination method provided in the above method embodiments.

[0222] The embodiments of the present application also provide a computer program product or a computer program, which comprises computer instructions stored in a computer readable storage medium. The processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device performs to implement the attribution index determination method provided in the above method embodiments.

[0223] Optionally, in the embodiments of the present application, the storage medium can be located in at least one network server of a plurality of network servers of a computer network. Optionally, in the embodiments, the storage medium can include but is not limited to various storage media that can store program codes, such as a U disk, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0224] The memory of the embodiments of the present application can be used to store software programs and modules. The processor performs various functions and data processing by running the software programs and modules stored in the memory. The memory can mainly include a program storage area and a data storage area. The program storage area can store operating systems, application programs required for functions, etc. The data storage area can store data created according to the use of the device, etc. In addition, the memory can include a high-speed random access memory, and can further include a non-volatile memory such as at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory device. Accordingly, the memory can further include a memory controller to provide the processor with access to the memory.

[0225] The attribution index determination method provided by the embodiments of the present application can be executed in a mobile terminal, a computer terminal, a server, or a similar computing device. Taking the case of running on a server as an example, Figure 19 is a hardware structure block diagram of a server of an attribution index determination method provided by the embodiments of the present application. As Figure 19 shown, the server 1900 can have great differences due to different configurations or performances, and can include one or more central processing units (CPU) 1910 (the processor 1910 can include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 1930 for storing data, one or more storage media 1920 (such as one or more mass storage devices) for storing application programs 1923 or data 1922. Among them, the memory 1930 and the storage medium 1920 can be temporary storage or persistent storage. The programs stored in the storage medium 1920 can include one or more modules, each of which can include a series of instruction operations in the server. Further, the central processing unit 1910 can be configured to communicate with the storage medium 1920 and execute a series of instruction operations in the storage medium 1920 on the server 1900. The server 1900 can also include one or more power supplies 1960, one or more wired or wireless network interfaces 1950, one or more input / output interfaces 1940, and / or one or more operating systems 1921, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

[0226] The input / output interface 1940 can be configured to receive or transmit data via a network. Examples of the network can include a wireless network provided by a communication provider of the server 1900. In an example, the input / output interface 1940 includes a network interface controller (NIC) that can be connected to other network devices through a base station to communicate with the Internet. In an example, the input / output interface 1940 can be a radio frequency (RF) module for communicating with the Internet in a wireless manner.

[0227] Those skilled in the art can understand that, Figure 19 The structure shown is only schematic, and does not limit the structure of the electronic device described above. For example, the server 1900 can further include more or fewer components than those shown, or have a different configuration of components than those shown. Figure 19 The structure shown is only schematic, and does not limit the structure of the electronic device described above. For example, the server 1900 can further include more or fewer components than those shown, or have a different configuration of components than those shown. Figure 19 The structure shown is only schematic, and does not limit the structure of the electronic device described above. For example, the server 1900 can further include more or fewer components than those shown, or have a different configuration of components than those shown.

[0228] From the above, it can be seen that the embodiments of the attribution index determination method, device, server, and storage medium provided by the present application can determine a target multimedia service index in a preset period of time that is in a downward trend, obtain a plurality of multimedia associated indexes corresponding to the target multimedia service index, construct an index logical tree according to the association relationship between the target multimedia service index and the multimedia associated indexes and the association relationship between the multimedia associated indexes, determine a target node that has a greater impact on a root node according to the association index between each parent node and the corresponding child node in the index logical tree, and determine a target multimedia associated index represented by the target node as an attribution index of the target multimedia service index, thereby quickly and accurately determining the attribution index of the target multimedia service index.

[0229] It should be noted that the above-mentioned order of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.

[0230] The various embodiments described in the specification are progressive in nature, and each of the embodiments can be incorporated into the other embodiments, and the same or similar parts among the embodiments can be mutually referred to. In particular, the device, equipment, and storage medium embodiments are basically similar to the method embodiments, and thus the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments.

[0231] Those skilled in the art can understand that all or part of the steps of the above-mentioned embodiments can be completed by hardware, or a program instructing relevant hardware, and the program can be stored in a computer storage medium, such as a read-only memory, a magnetic disk, or an optical disk.

[0232] The above description is only the preferred embodiment of the present application, and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An attribution metric determination method, characterized by, The method comprises: determining a target multimedia service index of a multimedia object in a preset time period, the target multimedia service index being in a downward trend in the preset time period; obtaining at least two multimedia associated indexes corresponding to the target multimedia service index; based on the target multimedia service index and the at least two multimedia associated indexes, constructing an index logical tree; a root node of the index logical tree represents the target multimedia service index, nodes other than the root node in the index logical tree represent the at least two multimedia associated indexes, and branches of the index logical tree represent the association relationship between different indexes; taking the root node of the index logical tree as a current node; determining a child node corresponding to the current node, and determining a current association coefficient between the current node and the corresponding child node according to a trend change curve of the index represented by the current node and a mode distance between the trend change curve and a trend change curve of the index represented by the child node; taking the child node with the current association coefficient greater than a preset threshold as a new current node; if the current node has a corresponding child node, repeating the steps of determining the current association coefficient between the current node and the corresponding child node, and taking the child node with the current association coefficient greater than the preset threshold as a new current node; if the current node has no corresponding child node, determining the current node as a target node; taking the multimedia associated index represented by the target node as a target multimedia associated index, and determining the target multimedia associated index as an attribution index of the target multimedia service index.

2. The method of claim 1, wherein, The current node represents the target multimedia service index, and the determination of the current association coefficient between the current node and the corresponding child node comprises: based on a target multimedia service index data set corresponding to the target multimedia service index, constructing a trend change curve of the target multimedia service index; based on a first multimedia associated index data set corresponding to the corresponding child node, constructing a trend change curve of the first multimedia associated index; determining a first mode distance between the trend change curve of the target multimedia service index and the trend change curve of the first multimedia associated index; taking the first mode distance as the association coefficient between the current node and the corresponding child node.

3. The method of claim 1, wherein, The current node represents a second multimedia associated index, and the determination of the current association coefficient between the current node and the corresponding child node comprises: based on a second multimedia associated index data set corresponding to the second multimedia associated index, constructing a trend change curve of the second multimedia associated index; based on a third multimedia associated index data set corresponding to the corresponding child node, constructing a trend change curve of the third multimedia associated index; determining a second mode distance between the trend change curve of the second multimedia associated index and the trend change curve of the third multimedia associated index; taking the second mode distance as the association coefficient between the current node and the corresponding child node.

4. The method of claim 1, wherein, The target nodes are at least two, the multimedia association index represented by the target nodes is taken as a target multimedia association index, and the target multimedia association index is determined as an attribution index of the target multimedia service index, including: Taking the multimedia association index represented by each target node as a target multimedia association index; Obtaining a historical index data set of each target multimedia association index in a historical period; Obtaining a current index data set of each target multimedia association index in the preset period; Determining, as a multimedia association index to be screened out, a target multimedia association index with a data change trend consistent with data in the historical index data set and data in the current index data set; Determining, as the attribution index of the target multimedia service index, an index other than the multimedia association index to be screened out in the at least two target multimedia association indexes.

5. The method according to any one of claims 1 to 4, characterized in that, The method for determining the target multimedia service index of the multimedia object in the preset period includes: Obtaining at least two multimedia service indexes of the multimedia object in the preset period; Determining, by a trend test method, an initial change trend of each multimedia service index in the preset period; Determining, as a candidate multimedia service index, a multimedia service index with a downward trend as the initial change trend; Determining, by a least square method, a slope of a trend change curve of the candidate multimedia service index; Determining, as the target multimedia service index, a candidate multimedia service index with a slope less than a preset slope threshold.

6. The method of claim 5, wherein, After the at least two multimedia service indexes of the multimedia object in the preset period are obtained, the method further includes: Obtaining a multimedia service index data set corresponding to each multimedia service index; Determining abnormal data in each multimedia service index data set, the abnormal data being data in a target period; Deleting the abnormal data in each multimedia service index data set to obtain an updated multimedia service index data set; Correspondingly, the method for determining, by the trend test method, the initial change trend of each multimedia service index in the preset period includes: Determining, by the trend test method, an initial change trend of data in each updated multimedia service index data set in the preset period.

7. An attribution index determination apparatus characterized by comprising: The device includes: A target multimedia service index determination module configured to determine a target multimedia service index of a multimedia object in a preset period, the target multimedia service index being in a downward trend in the preset period; A multimedia association index obtaining module configured to obtain at least two multimedia association indexes corresponding to the target multimedia service index; An index logical tree construction module configured to construct an index logical tree based on the target multimedia service index and the at least two multimedia association indexes, a root node of the index logical tree representing the target multimedia service index, nodes other than the root node in the index logical tree representing the at least two multimedia association indexes, and branches of the index logical tree representing association relationships between different indexes; and An index logical tree determination module configured to determine, by a trend test method, an initial change trend of each updated multimedia service index data set in the preset period. The target node determination module is configured to determine a target node based on a correlation coefficient between each parent node and a corresponding child node in the index logical tree, where the correlation coefficient represents an influence degree of the corresponding child node on the parent node; The attribution index determination module is configured to determine a multimedia correlation index represented by the target node as a target multimedia correlation index, and determine the target multimedia correlation index as an attribution index of the target multimedia service index. The target node determination module includes: a current node determination unit configured to determine a root node of the index logical tree as a current node; a current correlation coefficient determination unit configured to determine a child node corresponding to the current node, and determine a current correlation coefficient between the current node and the corresponding child node according to a trend change curve of an index represented by the current node and a mode distance between a trend change curve of an index represented by the child node; a current node updating unit configured to determine the child node with a current correlation coefficient greater than a preset threshold as a new current node; a step repeating unit configured to repeat the following steps if the current node has a corresponding child node: determining a current correlation coefficient between the current node and the corresponding child node; and determining the child node with a current correlation coefficient greater than a preset threshold as a new current node; and a target node determination unit configured to determine the current node as the target node if the current node has no corresponding child node.

8. The apparatus of claim 7, wherein, The current node represents the target multimedia service index, and the current correlation coefficient determination unit includes: a target multimedia service index trend change curve construction subunit configured to construct a trend change curve of the target multimedia service index based on a target multimedia service index data set corresponding to the target multimedia service index; a first curve construction subunit configured to construct a trend change curve of a first multimedia correlation index based on a first multimedia correlation index data set corresponding to the corresponding child node; a first mode distance determination subunit configured to determine a first mode distance between the trend change curve of the target multimedia service index and the trend change curve of the first multimedia correlation index; a first correlation coefficient determination subunit configured to determine the first mode distance as the correlation coefficient between the current node and the corresponding child node.

9. The apparatus of claim 7, wherein, The current node represents a second multimedia correlation index, and the current correlation coefficient determination unit includes: a second curve construction subunit configured to construct a trend change curve of the second multimedia correlation index based on a second multimedia correlation index data set corresponding to the second multimedia correlation index; a third curve construction subunit configured to construct a trend change curve of a third multimedia correlation index based on a third multimedia correlation index data set corresponding to the corresponding child node; a second mode distance determination subunit configured to determine a second mode distance between the trend change curve of the second multimedia correlation index and the trend change curve of the third multimedia correlation index; a second correlation coefficient determination subunit configured to determine the second mode distance as the correlation coefficient between the current node and the corresponding child node.

10. The apparatus of claim 7, wherein, The target nodes are at least two, and the attribution index determination module comprises: A target multimedia association index determination unit is configured to determine a multimedia association index represented by each target node as a target multimedia association index; A historical index dataset acquisition unit is configured to acquire a historical index dataset of each target multimedia association index in a historical period; A current index dataset unit is configured to acquire a current index dataset of each target multimedia association index in the preset period; A multimedia association index to be screened out determination unit is configured to determine a target multimedia association index with a consistent data change trend in the historical index dataset and the current index dataset as a multimedia association index to be screened out; An attribution index determination unit is configured to determine an index other than the multimedia association index to be screened out from the at least two target multimedia association indexes as an attribution index of the target multimedia service index.

11. The device according to any of claims 7-10, characterized in that The target multimedia service index determination module comprises: A multimedia service index acquisition unit is configured to acquire at least two multimedia service indexes of the multimedia object in the preset period; An initial change trend determination unit is configured to determine an initial change trend of each multimedia service index in the preset period based on a trend test method; A candidate multimedia service index determination unit is configured to determine a multimedia service index with a downward trend as a candidate multimedia service index; A slope determination unit is configured to determine a slope of a trend change curve of the candidate multimedia service index based on a least square method; A target multimedia service index determination unit is configured to determine a candidate multimedia service index with a slope less than a preset slope threshold as the target multimedia service index.

12. The apparatus of claim 11, wherein, The device further comprises: A multimedia service index dataset acquisition module is configured to acquire a multimedia service index dataset corresponding to each multimedia service index; An abnormal data determination module is configured to determine abnormal data in each multimedia service index dataset, the abnormal data being data in a target period; An abnormal data deletion module is configured to delete the abnormal data in each multimedia service index dataset to obtain an updated multimedia service index dataset; The initial change trend determination unit comprises: An initial change trend determination subunit is configured to determine an initial change trend of data in each updated multimedia service index dataset in the preset period based on the trend test method.

13. An attribution index determination device, characterized by, The device comprises a processor and a memory, the memory storing at least one instruction or at least one program, the at least one instruction or the at least one program being loaded and executed by the processor to implement the attribution index determination method according to any one of claims 1-6.

14. A computer storage medium, characterized in that The computer storage medium stores at least one instruction or at least one program, the at least one instruction or the at least one program being loaded and executed by a processor to implement the attribution index determination method according to any one of claims 1-6.

15. A computer program product, characterised in that, The computer program product comprises computer instructions, which, when executed by a processor, implement the attribution index determination method according to any one of claims 1-6.

Images