A data fluctuation attribution method and device, electronic equipment and storage medium

Abstract

Embodiments of the present disclosure disclose a data fluctuation attribution method and device, electronic equipment and storage medium, comprising: in response to the index data generating a preset fluctuation, obtaining at least one dimension related to the index data; each dimension contains at least one dimension item; determining the contribution rate and data volume proportion of each dimension item; the contribution rate is the influence parameter of the change amount of the business data corresponding to the index data under each dimension item on the index data; the data volume proportion is the proportion of the data volume belonging to each dimension item in the total data volume in the business data corresponding to the index data; according to the contribution rate of each dimension item, the contribution rate proportion of each dimension item under each dimension is determined; according to the contribution rate proportion and the data volume proportion of each dimension item under each dimension, the root cause dimension is determined from each dimension, and the root cause dimension item is determined from the root cause dimension. The traditional threshold-based attribution method is abandoned, and the attribution accuracy can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and more particularly to a method, apparatus, electronic device, and storage medium for attributing data fluctuations. Background Technology

[0002] In the internet environment, many business scenarios require attribution analysis of fluctuations in core metrics. Attribution analysis refers to the process of identifying the dimensions and items causing the fluctuations from among the various dimensions related to the metrics. The identified dimensions and items can be called the root cause dimension and root cause dimension item, respectively. By identifying the root cause dimension and root cause dimension item, targeted measures can be taken to ensure the normal operation of the business.

[0003] In existing technologies, the Adtributor algorithm is commonly used for automated attribution. This algorithm relies on two thresholds to filter dimensions and dimensional items that contribute significantly to volatility. However, existing technologies often suffer from unreasonable threshold settings, which can greatly affect the accuracy of attribution. Summary of the Invention

[0004] This disclosure provides a method, apparatus, electronic device, and storage medium for attributing data fluctuations, which can improve attribution accuracy.

[0005] In a first aspect, embodiments of this disclosure provide a method for attributing data fluctuations, including:

[0006] In response to a preset fluctuation in the indicator data, at least one dimension related to the indicator data is obtained; each dimension contains at least one dimension item.

[0007] Determine the contribution rate and data volume ratio of each dimension item; wherein, the contribution rate is the influence parameter of the change in the business data corresponding to the indicator data under each dimension item on the indicator data; the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume;

[0008] Based on the contribution rate of each dimension item, determine the contribution rate percentage of each dimension item under each dimension;

[0009] Based on the contribution rate and data volume of each dimension item under each dimension, the root cause dimension is determined from each dimension, and the root cause dimension item is determined from the root cause dimension.

[0010] Secondly, embodiments of this disclosure also provide an attribution device for data fluctuations, comprising:

[0011] The acquisition module is used to acquire at least one dimension related to the indicator data in response to a preset fluctuation in the indicator data; each dimension includes at least one dimension item;

[0012] The dimension item data determination module is used to determine the contribution rate and data volume ratio of each dimension item; wherein, the contribution rate is the influence parameter of the change in the business data corresponding to the indicator data under each dimension item on the indicator data; and the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume.

[0013] The contribution rate percentage determination module is used to determine the contribution rate percentage of each dimension item under each dimension based on the contribution rate of each dimension item.

[0014] The attribution module is used to determine the root cause dimension from each dimension based on the contribution rate and data volume ratio of each dimension item under each dimension, and to determine the root cause dimension item from the root cause dimension.

[0015] Thirdly, embodiments of this disclosure also provide an electronic device, the electronic device comprising:

[0016] One or more processors;

[0017] Storage device for storing one or more programs.

[0018] When the one or more programs are executed by the one or more processors, the one or more processors implement the data fluctuation attribution method as described in any embodiment of this disclosure.

[0019] Fourthly, embodiments of this disclosure also provide a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform an attribution method for data fluctuations as described in any of the embodiments of this disclosure.

[0020] In the technical solution of this disclosure embodiment, in response to a preset fluctuation in indicator data, at least one dimension related to the indicator data can be obtained; each dimension includes at least one dimension item; the contribution rate and data volume ratio of each dimension item are determined; wherein, the contribution rate is the influence parameter of the change in the business data corresponding to the indicator data under each dimension item on the indicator data; the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume; based on the contribution rate of each dimension item, the contribution rate ratio of each dimension item under each dimension is determined; based on the contribution rate ratio and data volume ratio of each dimension item under each dimension, the root cause dimension is determined from each dimension, and the root cause dimension item is determined from the root cause dimension.

[0021] In implementing the attribution method proposed in this disclosure, it was found that when indicator data fluctuates, there is a strong correlation between the contribution rate and data volume ratio of each dimension item included in non-root cause dimensions. For example, when the data volume ratio of a certain dimension item is high, its contribution rate ratio will also be high. Therefore, the root cause dimension can be located by examining the correlation between the contribution rate ratio and data volume ratio of each dimension item.

[0022] It was also found that when the indicator data fluctuated, the contribution rate and data volume of each non-root cause dimension under the root cause dimension also followed certain patterns. For example, the contribution rate and data volume of each non-root cause dimension were basically consistent. Therefore, the root cause dimension can also be located by examining the numerical patterns of the contribution rate and data volume of each dimension under the root cause dimension.

[0023] The attribution method provided in this embodiment can abandon the traditional threshold-based attribution method, which can avoid the situation where unreasonable threshold settings affect the accuracy of attribution and improve the accuracy of attribution. Attached Figure Description

[0024] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0025] Figure 1 A flowchart illustrating a data fluctuation attribution method provided in an embodiment of this disclosure;

[0026] Figure 2 A flowchart illustrating a data fluctuation attribution method provided in an embodiment of this disclosure;

[0027] Figure 3 A flowchart illustrating a data fluctuation attribution method provided in an embodiment of this disclosure;

[0028] Figure 4 A schematic diagram of the structure of a data fluctuation attribution device provided in an embodiment of this disclosure;

[0029] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0030] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0031] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0032] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0033] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0034] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0035] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0036] Figure 1 This is a flowchart illustrating a data fluctuation attribution method provided in an embodiment of this disclosure. This embodiment is applicable to situations where the dimensions related to the indicator data are stable, and attribution analysis is performed on the fluctuations in the indicator data. The method can be executed by a data fluctuation attribution device, which can be implemented in software and / or hardware, and can be configured in an electronic device, such as a computer device.

[0037] like Figure 1 As shown, the data fluctuation attribution method provided in this embodiment may include:

[0038] S110. In response to a preset fluctuation in the indicator data, obtain at least one dimension related to the indicator data; each dimension contains at least one dimension item.

[0039] In this embodiment of the disclosure, the indicator data can be used to characterize the operational status of the business, such as normal or abnormal status. The indicator data used may differ in different business scenarios. For example, the indicator data used in a live streaming scenario may include the success rate of stream acquisition, while the indicator data used in an advertising push scenario may include the access success rate, etc. At least one indicator data can be used in each business scenario to characterize whether the business is operating normally from different aspects.

[0040] Preset fluctuations in indicator data refer to situations where the change in indicator data exceeds a preset threshold within a unit of time, i.e., a sudden increase or decrease in the indicator data. The unit of time and the preset threshold can be set according to the specific business scenario.

[0041] Dimensions related to metric data can be considered as factors that can influence the magnitude of the metric data. Dimension items can be considered as optional items within a dimension. For example, for the metric data "success rate of attracting viewers" in a live streaming scenario, "province" can be a dimension affecting the magnitude of "success rate of attracting viewers," and "Province A" and "Province B" can be various dimension items under the "province" dimension.

[0042] In this embodiment, the dimensions related to the indicator data and the dimension items under each dimension are relatively stable, that is, they can be unchanging or their changes follow a certain pattern. For each business scenario, at least one dimension related to each indicator data and at least one dimension item under each dimension can be pre-set. In response to a preset fluctuation in a certain indicator data, the dimensions related to that indicator data and the dimension items under each dimension can be obtained according to the pre-set parameters to facilitate the initiation of subsequent attribution analysis steps.

[0043] S120. Determine the contribution rate and data volume ratio of each dimension item.

[0044] The contribution rate (explanatory power, EP) is the parameter representing the impact of changes in the corresponding business data across various dimensions on the metric data. The algorithm for calculating the contribution rate varies depending on the type of metric data and the corresponding business data. Metric data can include, but is not limited to, additive, divisional, and composite types. Additive metric data can include count-type metrics such as page views and sales volume; divisional metric data can include probability-based metrics such as traffic success rate and access success rate; composite metric data can be generated by combining multiple additive, divisional, or other basic metric data.

[0045] For additive metric data, the corresponding business data can be considered the metric data. For divisible metric data, the business data includes the business data corresponding to the numerator and the business data corresponding to the denominator. For example, for the metric data "stream success rate," the business data corresponding to the numerator can be the number of successful stream pulls, and the business data corresponding to the denominator can be the total number of stream pulls. The business data corresponding to the numerator can be simply referred to as the numerator data, and the business data corresponding to the denominator can be simply referred to as the denominator data.

[0046] For additive indicator data, the contribution rate can be calculated as follows: the change in business data under each dimension item, divided by the overall change in business data. For divisive indicator data, the contribution rate can be calculated as follows: using the indicator data to generate the overall numerator data before the preset change, add the change in the numerator data of each dimension item to obtain a first value; using the indicator data to generate the overall denominator data before the preset change, add the change in the denominator data of each dimension item to obtain a second value; dividing the first value by the second value to obtain the indicator data after the dimension item change; using the indicator data to generate the overall numerator data before the preset change, divide by the indicator data to generate the overall denominator data before the preset change to obtain the original indicator data; subtracting the original indicator data from the indicator data after the dimension item change to obtain the contribution rate.

[0047] For other types of indicator data, the contribution rate calculation method can refer to the contribution rate calculation method in the Adtributor algorithm. The specific business data corresponding to each indicator data is related to the business scenario, and will not be listed exhaustively here.

[0048] The data volume percentage can be defined as the proportion of data belonging to each dimension item in the business data corresponding to the indicator data out of the total data volume. The data volume percentage can be understood as the proportion of the size of each dimension item within each dimension. The sum of the data volume percentages of each dimension item within each dimension is 100%. Business data is typically dynamic, and the data volume percentage can be determined as follows: Multiple samples of business data can be taken; the proportion of data belonging to each dimension item in each sample can be calculated out of the total data volume; based on the proportions calculated each time, the data volume percentage of each dimension item can be estimated.

[0049] S130. Based on the contribution rate of each dimension item, determine the contribution rate percentage of each dimension item under each dimension.

[0050] The contribution rate of each dimension item under each dimension can represent the degree of contribution of each dimension item to the fluctuation of the indicator data under its respective dimension.

[0051] For additive indicator data, the sum of the contribution rates of all items within each dimension is 100%. When the contribution rate of each item is greater than 0, the contribution rate of each item within each dimension can represent the degree of contribution of each item to the fluctuation of the indicator data within its respective dimension. In this case, the contribution rate of each item can be directly used as the percentage of contribution of each item within each dimension.

[0052] However, for additive indicator data, the interpretability of the contribution rate is poor when the contribution rate of each dimension is less than or equal to 0. Furthermore, the interpretability of contribution rate values ​​is similarly poor for other types of indicator data, making it difficult for non-specialist personnel to determine the degree of contribution based on the numerical value of the contribution rate. In such cases, existing dimensionless methods can be used to determine the contribution percentage of each dimension under each dimension based on the numerical value of the contribution rate of each dimension. Compared to contribution rate alone, determining the contribution percentage of each dimension under each dimension can improve the interpretability of the attribution process.

[0053] In some optional implementations, determining the contribution rate percentage of each dimension item under each dimension based on the contribution rate of each dimension item can include: normalizing the contribution rate of each dimension item under each dimension to obtain the contribution rate percentage of each dimension item under each dimension. Among these optional implementations, a normalization algorithm from the dimensionless algorithm can be used to normalize the contribution rate of each dimension item under each dimension, ensuring that the contribution rate percentage of all dimension items under each dimension is greater than 0 and sums to 100%, thereby improving the interpretability of the attribution analysis process.

[0054] S140. Based on the contribution rate and data volume of each dimension item under each dimension, determine the root cause dimension from each dimension, and determine the root cause dimension item from the root cause dimension.

[0055] In implementing the attribution method proposed in this disclosure, it was found that when indicator data fluctuates, there is a strong correlation between the contribution rate and data volume of each dimension item within a non-root cause dimension. This can be understood as follows: the larger the proportion of a dimension item's size within a given dimension, the greater its contribution to the fluctuation of the indicator data should be.

[0056] In this embodiment, the root cause dimension can be identified by examining the correlation between the contribution rate percentage and the data volume percentage of each dimension item. For example, when the distribution of the contribution rate percentage and the data volume percentage of each dimension item are basically consistent, the dimension can be considered a non-root cause dimension; when the distribution of the contribution rate percentage and the data volume percentage of each dimension item differs significantly, the dimension can be considered a root cause dimension. The method for determining the root cause dimension provided in this embodiment abandons the traditional method of using fixed thresholds to screen root cause dimensions, thereby improving the accuracy of dimensional attribution.

[0057] Furthermore, it was found that when the indicator data fluctuated, the contribution rate and data volume of each non-root cause dimension under the root cause dimension also followed certain patterns. For example, the contribution rate and data volume of each non-root cause dimension were basically consistent.

[0058] In this embodiment of the disclosure, the root cause dimension item can be located by examining the numerical patterns of the contribution rate and data volume ratio of each dimension item under the root cause dimension. For example, under the root cause dimension, when the contribution rate ratio and data volume ratio of a certain dimension item are basically the same, the dimension item can be considered a non-root cause dimension item; when the contribution rate ratio and data volume ratio of a certain dimension item differ significantly, the dimension item can be considered a root cause dimension item. The method for determining the root cause dimension item provided in this embodiment of the disclosure abandons the traditional method of using fixed thresholds to screen root cause dimension items, which can improve the attribution accuracy of dimension items.

[0059] The technical solution of this disclosure embodiment, in response to a preset fluctuation in indicator data, obtains dimensions related to the indicator data; each dimension includes at least one dimension item; determines the contribution rate and data volume ratio of each dimension item; wherein, the contribution rate is the influence parameter of the change in the business data corresponding to the indicator data under each dimension item on the indicator data; the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume; based on the contribution rate of each dimension item, determines the contribution rate ratio of each dimension item under each dimension; based on the contribution rate ratio and data volume ratio of each dimension item under each dimension, determines the root cause dimension from each dimension, and determines the root cause dimension item from the root cause dimension.

[0060] By analyzing the contribution rate and data volume of each dimension, the root cause dimension and its specific items can be identified. This approach abandons the traditional threshold-based attribution method, avoiding situations where inappropriate threshold settings negatively impact attribution accuracy and thus improving overall accuracy.

[0061] This embodiment can be combined with various optional solutions in the data fluctuation attribution method provided in the above embodiments. The data fluctuation attribution method provided in this embodiment describes the root cause dimension and root cause dimension items in detail. The root cause dimension can be located by the similarity of the sequence distribution between the contribution rate percentage and the data volume percentage of each dimension item. The root cause dimension can also be located by the deviation between the contribution rate percentage and the data volume percentage of each dimension item.

[0062] Figure 2 This is a flowchart illustrating a data fluctuation attribution method provided in an embodiment of this disclosure. Figure 2 As shown, the data fluctuation attribution method provided in this embodiment may include:

[0063] S210. In response to a preset fluctuation in the indicator data, obtain at least one dimension related to the indicator data; each dimension contains at least one dimension item.

[0064] S220. Determine the contribution rate and data volume ratio of each dimension item.

[0065] Among them, the contribution rate is the parameter that measures the change in the business data corresponding to the indicator data under each dimension. The data volume ratio is the proportion of the data volume belonging to each dimension in the business data corresponding to the indicator data to the total data volume.

[0066] S230. Based on the contribution rate of each dimension item, determine the contribution rate percentage of each dimension item under each dimension.

[0067] S240. Take the sequence of contribution rate percentage of each dimension item under each dimension as the first sequence, and the sequence of data volume percentage of each dimension item as the second sequence.

[0068] The first sequence represents the distribution of the contribution rate of each dimension item under the corresponding dimension, while the second sequence represents the distribution of the data volume of each dimension item under the corresponding dimension.

[0069] S250. Based on the similarity between the first sequence and the second sequence under each dimension, determine the root cause dimension from each dimension.

[0070] The similarity between the first and second sequences can be determined using existing algorithms for measuring the differences between distributions. A higher similarity between the first and second sequences indicates a lower probability that this dimension is a root cause dimension.

[0071] In some alternative implementations, similarity can be determined based on the following steps: determining the similarity between the first sequence and the second sequence according to the distribution distance in each dimension; wherein the distribution distance is inversely correlated with the similarity.

[0072] The distributional distance between the first and second sequences can be determined by calculating at least one of the following: JS divergence, KL divergence, Wasserstein distance, and cross-entropy. JS divergence, KL divergence, Wasserstein distance, and cross-entropy are positively correlated with the distributional distance, while the distributional distance is inversely correlated with similarity. For example, the larger the JS divergence between the first and second sequences, the larger the distributional distance, and the lower their similarity. Furthermore, other algorithms for determining the distributional distance between the first and second sequences can also be applied here, but are not exhaustively listed.

[0073] In some optional implementations, the root cause dimension is determined from each dimension based on the similarity between the first and second sequences under each dimension. This can include selecting a predetermined number of dimensions with low similarity as the root cause dimensions. In these optional implementations, the dimensions can be sorted according to their similarity, for example, in descending or ascending order. If the dimensions are sorted in descending order, the predetermined number of dimensions with low similarity can be considered the last predetermined number of dimensions; if the dimensions are sorted in ascending order, the predetermined number of dimensions with low similarity can be considered the first predetermined number of dimensions.

[0074] Among these, the lower the similarity, the greater the likelihood that the dimension is the root cause. For example, when sorting the dimensions in descending order of JS divergence (i.e., sorting them in ascending order of similarity), the earlier a dimension appears in the sorted list, the lower the similarity between its first and second sequences, and the greater the likelihood that it is the root cause dimension. In this case, a predetermined number of dimensions (e.g., the top 3, top 5, etc.) can be selected as potential root cause dimensions.

[0075] S260. Based on the deviation between the contribution rate and data volume of each dimension item under the root cause dimension, determine the root cause dimension items from each dimension item under the root cause dimension.

[0076] In this embodiment, the difference between the contribution rate and data volume ratio of each dimension item under the root cause dimension can be calculated. If the difference is small or non-existent, the dimension item is considered to be unlikely to be a root cause dimension item.

[0077] In some optional implementations, the root cause dimension item is determined from the various dimension items under the root cause dimension based on the deviation between the contribution rate percentage and the data volume percentage of each dimension item. This can include: obtaining the correspondence between the data volume percentage and the allowable deviation value; determining the target deviation value corresponding to the data volume percentage of each dimension item under the root cause dimension based on the correspondence; comparing the deviation between the contribution rate percentage and the data volume percentage of each dimension item with the corresponding target deviation value; and determining the root cause dimension item from the various dimension items based on the comparison result.

[0078] In these optional implementation methods, the allowable deviation values ​​corresponding to different data volume proportions may differ, and the correspondence between the two can be pre-set based on empirical or experimental values. This correspondence can be pre-set using tables, functions, or other methods. Then, based on the pre-set correspondence, the allowable deviation value corresponding to the current data volume proportion of each dimension item, i.e., the target deviation value, can be determined. When the deviation between the contribution rate proportion of a dimension item and its data volume proportion is less than the target deviation value, the dimension item can be considered a non-root cause dimension item; when the deviation between the contribution rate proportion of a dimension item and its data volume proportion is greater than or equal to the target deviation value, the dimension item can be considered a root cause dimension item.

[0079] In the technical solutions of this disclosure, the root cause dimension and root cause dimension items are described in detail. The root cause dimension can be located by the similarity of the sequence distribution between the contribution rate percentage and the data volume percentage of each dimension item. The root cause dimension can also be located by the deviation between the contribution rate percentage and the data volume percentage of each dimension item. The data fluctuation attribution method provided in this disclosure belongs to the same disclosed concept as the data fluctuation attribution method provided in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and the same technical features have the same beneficial effects in this embodiment and the above embodiments.

[0080] This disclosure embodiment can be combined with various optional solutions in the data fluctuation attribution methods provided in the above embodiments. The data fluctuation attribution method provided in this embodiment details the attribution method for indicator data fluctuations in live streaming scenarios. By considering the contribution rate and data volume ratio of each dimension item, attribution analysis can be performed on the fluctuations of various indicator data in live streaming scenarios.

[0081] Figure 3 This is a flowchart illustrating a data fluctuation attribution method provided in an embodiment of this disclosure. Figure 3 As shown, the data fluctuation attribution method provided in this embodiment, applied to live streaming servers, may include:

[0082] S301. In response to a preset fluctuation in the indicator data, obtain at least one dimension related to the indicator data; each dimension contains at least one dimension item.

[0083] In live streaming scenarios, the metrics data include at least one of the following: streaming success rate, duration of buffering per unit time, number of buffering events per unit time, latency, and first frame duration.

[0084] The success rate of streaming access represents the proportion of users who successfully access the stream; a higher success rate is better. The duration of buffering and the number of buffering events per unit of time (e.g., duration of buffering per 100 seconds and number of buffering events per 100 seconds) directly describe the smoothness of the live stream viewed by the user. The latency can be understood as end-to-end latency, also known as the hand-waving latency, which is the time between the broadcaster's recorded hand-waving action and the user's playback of that action; a shorter latency is better. The first frame duration is the time from clicking on the live stream to seeing the first frame; a shorter first frame duration is better.

[0085] Dimensions related to the aforementioned metrics data may include, for example, live stream parameters, content delivery network, application version number, province, carrier, operating system, etc. In this embodiment, the dimension can be represented as D = {D1, D2, ..., D...} m} indicates that for any dimension D i (1≤i≤m), can include N dimension items, and N is a positive integer.

[0086] In this embodiment, the base period can be understood as a time period when the indicator data is normal, and the analysis period can be understood as a time period when the indicator data is abnormal. When the indicator data exhibits a preset fluctuation, it can be considered that the current period has transitioned from the base period to the analysis period. By obtaining the dimensions related to the indicator data and the dimension items under each dimension, it is possible to pinpoint the reasons for the significant fluctuation in live broadcast quality during the analysis period compared to the base period.

[0087] S302. Determine the contribution rate and data volume ratio of each dimension item.

[0088] The contribution rate is the parameter representing the impact of changes in the business data corresponding to the indicator data across each dimension on the indicator data. For example, for division-type indicators such as traffic acquisition success rate, assume the overall market value is 'b' in the base period and 'a' in the analysis period. The j-th dimension item of the i-th dimension can be referred to as E. ij (1≤j≤n), its value in the base period is b ij The value during the analysis period is a ij Where b = b n *b d , a = a n / a d , The denominator represents the total number of pull-throughs, and the numerator represents the number of successful pull-throughs. E ij The contribution rate can be EP ij And it can be used with formulas express.

[0089] The data volume percentage refers to the proportion of data belonging to each dimension item in the business data corresponding to the indicator data to the total data volume. Specifically, for E... ij The proportion of data in the base period can be Percentage of data volume during the analysis period By combining the data volume proportions of the base period and the analysis period, we can fit the proportion of data belonging to each dimension in the actual sampled data to the total data volume. For example...

[0090] S303. Based on the contribution rate of each dimension item, determine the contribution rate percentage of each dimension item under each dimension.

[0091] This involves normalizing the contribution rate of each dimension's items within each dimension to obtain the percentage contribution of each dimension's items within each dimension. For example, for each dimension D... i E below ij When the minimum contribution rate is greater than or equal to 0, the contribution rate percentage is... It can be determined using the following formula:

[0092]

[0093] Where, N i It can represent the number of dimension items in the i-th dimension.

[0094] Otherwise, contribution rate percentage It can be determined using the following formula:

[0095]

[0096]

[0097] Among them, EP i max EP can represent the maximum contribution rate under the i-th dimension. i min It can represent the minimum contribution rate under the i-th dimension.

[0098] S304. Take the sequence of contribution rate percentage of each dimension item under each dimension as the first sequence, and the sequence of data volume percentage of each dimension item as the second sequence.

[0099] Wherein, the i-th dimension D i The sequence composed of the contribution rate percentages of each dimension can be used as EP. i This indicates that the sequence composed of the proportion of data in each dimension can be represented by P. i express.

[0100] S305. The similarity is determined based on the distribution distance between the first sequence and the second sequence under each dimension; where the distribution distance and similarity are inversely correlated.

[0101] Among them, it can be achieved through code segment D i .JSD=D JS (EP i ,P i To determine dimension D i Next first sequence EP i With the second sequence P i JS divergence.

[0102] S306. Among all dimensions, the pre-set number of dimensions with low similarity are taken as root cause dimensions.

[0103] Among them, it can be achieved through code segment D i `.SortDescend(JSD)` sorts the dimensions in descending order of JavaScript divergence. Furthermore, this can be achieved through code snippet D. i .Take(5) selects the first 5 dimensions as possible root cause dimensions.

[0104] S307. The correspondence between the proportion of acquired data and the allowable deviation value.

[0105] The relationship between the data volume ratio and the allowable deviation value can be set by using a piecewise function that presets the data volume ratio and the allowable deviation value.

[0106] For example, when the data volume percentage P is (0, 20), the allowable deviation can be within 0.75*P+5; when the data volume percentage P is [20, 60), the allowable deviation can be within 20; and when the data volume percentage P is [60, 100), the allowable deviation can be within -0.375*P+42.5. The piecewise function for the data volume percentage can be set based on empirical or experimental values, and is not limited to the examples described above.

[0107] S308. Based on the correspondence, determine the target deviation value corresponding to the data volume ratio of each dimension item under the root cause dimension.

[0108] Specifically, the target deviation value corresponding to the data volume ratio of each dimension item is determined based on a piecewise function of preset data volume ratio and allowable deviation value.

[0109] S309. Compare the deviation between the contribution rate and the data volume ratio of each dimension with the corresponding target deviation value.

[0110] Among them, it can be expressed by formula Calculate the deviation between the contribution rate percentage and the data volume percentage. It can also compare the magnitude of the diff with the target deviation value.

[0111] S310. Based on the comparison results, determine the root cause dimension items from each dimension item.

[0112] If the diff is greater than or equal to the target deviation value, the corresponding dimension can be identified as a root cause dimension item. Finally, the root cause dimension containing the root cause dimension items, as well as the root cause dimension items under each root cause dimension, can be output.

[0113] Experiments have verified that, in the attribution process of various indicator data in live streaming scenarios, the attribution method provided in this embodiment has a significant improvement in attribution accuracy in terms of root cause dimension and root cause dimension items compared to the traditional threshold-based attribution method.

[0114] This disclosure provides a detailed description of the attribution method for fluctuations in indicator data during live streaming. By analyzing the contribution rate and data volume of each dimension, attribution analysis can be performed on the fluctuations of various indicator data during live streaming. The attribution method for data fluctuations provided in this disclosure belongs to the same concept as the attribution method for data fluctuations provided in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and the same technical features have the same beneficial effects in this embodiment and the above embodiments.

[0115] Figure 4 This is a schematic diagram of a data fluctuation attribution device provided in an embodiment of this disclosure. The data fluctuation attribution device provided in this embodiment is applicable to situations where the dimensions related to the indicator data are stable, and attribution analysis is performed on the fluctuations of the indicator data.

[0116] like Figure 4 As shown, the data fluctuation attribution device provided in this embodiment of the disclosure may include:

[0117] The acquisition module 410 is used to acquire at least one dimension related to the indicator data in response to preset fluctuations in the indicator data; each dimension contains at least one dimension item;

[0118] The dimension item data determination module 420 is used to determine the contribution rate and data volume ratio of each dimension item; where the contribution rate is the parameter that affects the change of the business data corresponding to the indicator data under each dimension item on the indicator data; the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume.

[0119] The contribution rate percentage determination module 430 is used to determine the contribution rate percentage of each dimension item based on the contribution rate of each dimension item.

[0120] The attribution module 440 is used to determine the root cause dimension from each dimension and the root cause dimension item from each root cause dimension based on the contribution rate and data volume of each dimension item under each dimension.

[0121] In some optional implementations, the contribution rate determination module can be used for:

[0122] The contribution rate of each dimension item under each dimension is normalized to obtain the contribution rate percentage of each dimension item under each dimension.

[0123] In some alternative implementations, the attribution module can be used for:

[0124] The sequence consisting of the contribution rate of each dimension item under each dimension is taken as the first sequence, and the sequence consisting of the data volume of each dimension item is taken as the second sequence.

[0125] Based on the similarity between the first and second sequences under each dimension, the root cause dimension is determined from each dimension.

[0126] In some alternative implementations, the attribution module can be used for:

[0127] The dimensions with the lowest similarity and the highest number of similarities among all dimensions are selected as the root cause dimensions.

[0128] In some alternative implementations, the cloning module can determine similarity based on the following steps:

[0129] The similarity is determined by the distribution distance between the first sequence and the second sequence under each dimension; where the distribution distance and similarity are inversely correlated.

[0130] In some alternative implementations, the attribution module can be used for:

[0131] Based on the deviations in the contribution rate and data volume of each dimension item under the root cause dimension, the root cause dimension items are determined from the various dimension items under the root cause dimension.

[0132] In some alternative implementations, the attribution module can be used for:

[0133] The relationship between the proportion of acquired data and the allowable deviation value;

[0134] Based on the correspondence, determine the target deviation value corresponding to the data volume ratio of each dimension item under the root cause dimension;

[0135] Compare the deviation between the contribution rate and the data volume ratio of each dimension with the corresponding target deviation value.

[0136] Based on the comparison results, the root cause dimension items are determined from each dimension item.

[0137] In some optional implementations, the data fluctuation attribution device can be applied to live streaming servers; correspondingly, the metric data includes at least one of the following: streaming success rate, stuttering duration per unit time, stuttering frequency per unit time, latency duration, and first frame duration.

[0138] The data fluctuation attribution device provided in this disclosure can execute the data fluctuation attribution method provided in any embodiment of this disclosure, and has the corresponding functional modules and beneficial effects of the execution method.

[0139] It is worth noting that the various units and modules included in the above-mentioned device are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of each functional unit are only for easy differentiation and are not used to limit the protection scope of the embodiments of this disclosure.

[0140] The following is for reference. Figure 5 It illustrates an electronic device suitable for implementing embodiments of the present disclosure (e.g., Figure 5 The diagram below shows the structure of the terminal device or server 500. The terminal device in this embodiment may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), and vehicle terminals (e.g., vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0141] like Figure 5 As shown, the electronic device 500 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 502 or a program loaded from storage device 508 into random access memory (RAM) 503. The RAM 503 also stores various programs and data required for the operation of the electronic device 500. The processing unit 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0142] Typically, the following devices can be connected to I / O interface 505: input devices 506 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 507 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 508 including, for example, magnetic tapes, hard disks, etc.; and communication devices 509. Communication device 509 allows electronic device 500 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 5 An electronic device 500 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0143] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 509, or installed from a storage device 508, or installed from a ROM 502. When the computer program is executed by the processing device 501, it performs the functions defined in the data fluctuation attribution method of embodiments of this disclosure.

[0144] The electronic device provided in this embodiment and the data fluctuation attribution method provided in the above embodiments belong to the same disclosed concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0145] This disclosure provides a computer storage medium storing a computer program that, when executed by a processor, implements the data fluctuation attribution method provided in the above embodiments.

[0146] It should be noted that the computer-readable medium described in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor device or apparatus, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory (FLASH), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be executed by instructions, used by a device or apparatus, or used in conjunction with it. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such transmitted data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with instructions, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0147] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0148] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0149] The aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to:

[0150] In response to preset fluctuations in indicator data, at least one dimension related to the indicator data is obtained; each dimension contains at least one dimension item; the contribution rate and data volume ratio of each dimension item are determined; wherein, the contribution rate is the parameter that affects the indicator data by the change in the business data corresponding to the indicator data under each dimension item; the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume; based on the contribution rate of each dimension item, the contribution rate ratio of each dimension item under each dimension is determined; based on the contribution rate ratio and data volume ratio of each dimension item under each dimension, the root cause dimension is determined from each dimension, and the root cause dimension item is determined from the root cause dimension.

[0151] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including but not limited to object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0152] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using dedicated hardware-based implementations that perform the specified functions or operations, or using a combination of dedicated hardware and computer instructions.

[0153] The units described in the embodiments of this disclosure can be implemented in software or hardware. The names of the units and modules do not, in certain circumstances, constitute a limitation on the unit or module itself.

[0154] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Array (FPGA), Application Specific Integrated Circuit (ASIC), Application Specific Standard Parts (ASSP), System on Chip (SOC), Complex Programmable Logic Device (CPLD), and so on.

[0155] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in connection with the execution of instructions, an apparatus, or a device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor apparatus or device, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0156] According to one or more embodiments of this disclosure, [Example 1] provides an attribution method for data fluctuations, the method comprising:

[0157] In response to a preset fluctuation in the indicator data, at least one dimension related to the indicator data is obtained; each dimension contains at least one dimension item.

[0158] Determine the contribution rate and data volume ratio of each dimension item; wherein, the contribution rate is the influence parameter of the change in the business data corresponding to the indicator data under each dimension item on the indicator data; the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume;

[0159] Based on the contribution rate of each dimension item, determine the contribution rate percentage of each dimension item under each dimension;

[0160] Based on the contribution rate and data volume of each dimension item under each dimension, the root cause dimension is determined from each dimension, and the root cause dimension item is determined from the root cause dimension.

[0161] According to one or more embodiments of this disclosure, [Example 2] provides an attribution method for data fluctuations, further comprising:

[0162] In some optional implementations, determining the contribution rate percentage of each dimension item under each dimension based on the contribution rate of each dimension item includes:

[0163] The contribution rate of each dimension item under each dimension is normalized to obtain the contribution rate percentage of each dimension item under each dimension.

[0164] According to one or more embodiments of this disclosure, [Example 3] provides an attribution method for data fluctuations, further comprising:

[0165] In some optional implementations, determining the root cause dimension from each dimension based on the contribution rate and data volume ratio of each dimension item includes:

[0166] The sequence consisting of the contribution rate percentage of each dimension item under each dimension is taken as the first sequence, and the sequence consisting of the data volume percentage of each dimension item is taken as the second sequence.

[0167] The root cause dimension is determined from the dimensions based on the similarity between the first sequence and the second sequence under each dimension.

[0168] According to one or more embodiments of this disclosure, [Example 4] provides an attribution method for data fluctuations, further comprising:

[0169] In some optional implementations, determining the root cause dimension from the dimensions based on the similarity between the first sequence and the second sequence under each dimension includes:

[0170] Among all the dimensions, the preset number of dimensions with low similarity and the highest similarity are taken as root cause dimensions.

[0171] According to one or more embodiments of this disclosure, [Example 5] provides an attribution method for data fluctuations, further comprising:

[0172] In some alternative implementations, the similarity is determined based on the following steps:

[0173] The similarity is determined based on the distribution distance between the first sequence and the second sequence under each dimension; wherein the distribution distance is inversely correlated with the similarity.

[0174] According to one or more embodiments of this disclosure, [Example Six] provides an attribution method for data fluctuations, further comprising:

[0175] In some optional implementations, determining the root cause dimension items from the root cause dimensions includes:

[0176] Based on the deviation between the contribution rate and data volume ratio of each dimension item under the root cause dimension, the root cause dimension item is determined from each dimension item under the root cause dimension.

[0177] According to one or more embodiments of this disclosure, [Example Seven] provides an attribution method for data fluctuations, further comprising:

[0178] In some optional implementations, determining the root cause dimension item from the dimension items under the root cause dimension based on the deviation between the contribution rate and data volume ratio of each dimension item under the root cause dimension includes:

[0179] The relationship between the proportion of acquired data and the allowable deviation value;

[0180] Based on the correspondence, determine the target deviation value corresponding to the data volume ratio of each dimension item under the root cause dimension;

[0181] The deviation between the contribution rate percentage and the data volume percentage of each dimension item is compared with the corresponding target deviation value.

[0182] Based on the comparison results, the root cause dimension items are determined from each of the stated dimension items.

[0183] According to one or more embodiments of this disclosure, [Example Eight] provides an attribution method for data fluctuations, further comprising:

[0184] In some optional implementations, it is applied to live streaming servers; accordingly, the indicator data includes at least one of the following: streaming success rate, stuttering duration per unit time, stuttering frequency per unit time, latency duration, and first frame duration.

[0185] According to one or more embodiments of this disclosure, [Example Nine] provides a data fluctuation attribution device, the device comprising:

[0186] The acquisition module is used to acquire at least one dimension related to the indicator data in response to a preset fluctuation in the indicator data; each dimension includes at least one dimension item;

[0187] The dimension item data determination module is used to determine the contribution rate and data volume ratio of each dimension item; wherein, the contribution rate is the influence parameter of the change in the business data corresponding to the indicator data under each dimension item on the indicator data; and the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume.

[0188] The contribution rate percentage determination module is used to determine the contribution rate percentage of each dimension item under each dimension based on the contribution rate of each dimension item.

[0189] The attribution module is used to determine the root cause dimension from each dimension based on the contribution rate and data volume ratio of each dimension item under each dimension, and to determine the root cause dimension item from the root cause dimension.

[0190] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

[0191] Furthermore, while the operations are described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in a sequential order. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0192] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. An attribution method for data fluctuations, characterized in that, include: In response to a preset fluctuation in the indicator data, at least one dimension related to the indicator data is obtained; each dimension contains at least one dimension item. Determine the contribution rate and data volume ratio of each dimension item; wherein, the contribution rate is the influence parameter of the change in the business data corresponding to the indicator data under each dimension item on the indicator data; the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume; the data volume ratio is used to characterize the scale ratio of each dimension item under each dimension. Based on the contribution rate of each dimension item, determine the contribution rate percentage of each dimension item under each dimension; Based on the contribution rate and data volume of each dimension item under each dimension, the root cause dimension is determined from each dimension, and the root cause dimension item is determined from the root cause dimension. The step of determining the root cause dimension from each dimension based on the contribution rate and data volume percentage of each dimension item includes: The sequence consisting of the contribution rate percentage of each dimension item under each dimension is taken as the first sequence, and the sequence consisting of the data volume percentage of each dimension item is taken as the second sequence; wherein, the first sequence represents the distribution of the contribution rate percentage of each dimension item under the dimension, and the second sequence represents the distribution of the data volume percentage of each dimension item under the dimension. Based on the similarity between the first sequence and the second sequence under each of the dimensions, the root cause dimension is determined from each of the dimensions; wherein, the higher the similarity between the first sequence and the second sequence, the lower the probability that the corresponding dimension is the root cause dimension.

2. The method according to claim 1, characterized in that, The step of determining the contribution rate percentage of each dimension item under each dimension based on the contribution rate of each dimension item includes: The contribution rate of each dimension item under each dimension is normalized to obtain the contribution rate percentage of each dimension item under each dimension.

3. The method according to claim 1, characterized in that, The step of determining the root cause dimension from the dimensions based on the similarity between the first sequence and the second sequence under each dimension includes: Among all the dimensions, the preset number of dimensions with low similarity and the highest similarity are taken as root cause dimensions.

4. The method according to claim 1, characterized in that, The similarity is determined based on the following steps: The similarity is determined based on the distribution distance between the first sequence and the second sequence under each dimension; wherein the distribution distance is inversely correlated with the similarity.

5. The method according to claim 1, characterized in that, Determining the root cause dimension item from the root cause dimensions includes: Based on the deviation between the contribution rate and data volume ratio of each dimension item under the root cause dimension, the root cause dimension item is determined from each dimension item under the root cause dimension.

6. The method according to claim 5, characterized in that, The step of determining the root cause dimension item from each dimension item under the root cause dimension based on the deviation between the contribution rate percentage and the data volume percentage of each dimension item under the root cause dimension includes: The relationship between the proportion of acquired data and the allowable deviation value; Based on the correspondence, determine the target deviation value corresponding to the data volume ratio of each dimension item under the root cause dimension; The deviation between the contribution rate percentage and the data volume percentage of each dimension item is compared with the corresponding target deviation value. Based on the comparison results, the root cause dimension items are determined from each of the stated dimension items.

7. The method according to any one of claims 1-6, characterized in that, It is applied to live streaming servers; accordingly, the indicator data includes at least one of the following: streaming success rate, stuttering duration per unit time, stuttering frequency per unit time, latency duration, and first frame duration.

8. An attribution device for data fluctuations, characterized in that, include: The acquisition module is used to acquire at least one dimension related to the indicator data in response to a preset fluctuation in the indicator data; each dimension includes at least one dimension item. The dimension item data determination module is used to determine the contribution rate and data volume ratio of each dimension item; wherein, the contribution rate is the influence parameter of the change in the business data corresponding to the indicator data under each dimension item on the indicator data; the data volume ratio is the proportion of the data volume belonging to each dimension item in the business data corresponding to the indicator data to the total data volume; the data volume ratio is used to characterize the scale ratio of each dimension item under each dimension. The contribution rate percentage determination module is used to determine the contribution rate percentage of each dimension item under each dimension based on the contribution rate of each dimension item. The attribution module is used to determine the root cause dimension from each dimension based on the contribution rate and data volume ratio of each dimension item under each dimension, and to determine the root cause dimension item from the root cause dimension. The contribution rate determination module is specifically used for: The sequence consisting of the contribution rate percentage of each dimension item under each dimension is taken as the first sequence, and the sequence consisting of the data volume percentage of each dimension item is taken as the second sequence; wherein, the first sequence represents the distribution of the contribution rate percentage of each dimension item under the dimension, and the second sequence represents the distribution of the data volume percentage of each dimension item under the dimension. Based on the similarity between the first sequence and the second sequence under each of the dimensions, the root cause dimension is determined from each of the dimensions; wherein, the higher the similarity between the first sequence and the second sequence, the lower the probability that the corresponding dimension is the root cause dimension.

9. An electronic device, characterized in that, The electronic device includes: One or more processors; Storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the attribution method for data fluctuations as described in any one of claims 1-7.

10. A storage medium comprising computer-executable instructions, which, when executed by a computer processor, are used to perform the attribution method for data fluctuations as described in any one of claims 1-7.

Images