A method and device for storing user data

Abstract

The present invention discloses a method and device for storing user data, and relates to the field of computer technology. A specific implementation of the method includes: performing rank distribution conversion on the index value of each user, and calculating the rank ranking rate of each user; performing distribution mapping on the rank ranking rate of each user based on the inverse function of the cumulative distribution function, and obtaining the mapping result value of each user; determining the index performance level of each user based on the mapping threshold and the mapping result value of each user; storing the mapping result value and the index performance level of each user in a big data platform for database system call. This implementation can solve the technical problem that the data of each user stored in the big data platform is not accurate enough.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method and device for storing user data. Background Art

[0002] A large number of third-party merchants have opened stores on e-commerce platforms. The large amount of sales activities every day have brought a large amount of new data. E-commerce platforms need to judge the service performance level of third-party merchants based on the feedback of some data indicators to ensure that their platform users can enjoy high-quality services. In the process of judging the service performance level of third-party merchants, for a certain data indicator, a standardized method is currently mainly used to judge the indicator performance level.

[0003] In the process of implementing the present invention, the inventors found that there are at least the following problems in the prior art:

[0004] Since data indicators may have problems such as small samples, skewed indicator distribution or abnormal extreme values, the use of standardized methods to identify indicator performance levels will lead to inaccurate identification results, which in turn will lead to inaccurate data stored in the big data platform for each user, affecting the use of other business systems. Summary of the invention

[0005] In view of this, an embodiment of the present invention provides a method and apparatus for storing user data to solve the technical problem that the data of each user stored in a big data platform is not accurate enough.

[0006] To achieve the above object, according to one aspect of an embodiment of the present invention, a method for storing user data is provided, comprising:

[0007] Perform rank distribution conversion on the indicator value of each user, and calculate the rank rate of each user;

[0008] Based on the inverse function of the cumulative distribution function, the rank rate of each user is distributed and mapped to obtain the mapping result value of each user;

[0009] Determining the indicator performance level of each user based on the mapping threshold and the mapping result value of each user;

[0010] The mapping result values ​​and indicator performance levels of each user are stored in the big data platform for the database system to call.

[0011] Optionally, the index value of each user is transformed into a rank distribution, including:

[0012] The index values ​​of each user are ranked with the goal of having better index value performance and smaller rank.

[0013] Optionally, the index values ​​of each user are ranked with the goal of having better index values ​​and smaller rank, including:

[0014] If the indicator value shows that the larger the indicator value, the smaller the corresponding rank, then the users are sorted in descending order of the indicator value;

[0015] If the indicator value shows that the smaller the indicator value, the smaller the corresponding rank, then the users are sorted in order from small to large indicator values.

[0016] Optionally, for any user, the rank rate of the user is calculated using the following method:

[0017] The rank rate of the user is calculated according to the total number of users in the ranking and the rank of the user in the ranking.

[0018] Optionally, distribution mapping is performed on the rank rate of each user based on an inverse function of a cumulative distribution function, including:

[0019] If the distribution of the index values ​​of the respective users is a skewed distribution, then distribution mapping is performed on the rank rates of the respective users based on the inverse function of the cumulative distribution function of the exponential distribution;

[0020] If the distribution of the indicator values ​​of the various users is a normal distribution, then the rank rates of the various users are distributed and mapped based on the inverse function of the cumulative distribution function of the bell-shaped distribution.

[0021] Optionally, before performing rank distribution conversion on the indicator value of each user, the following steps are further included:

[0022] Determine whether there are users whose number of users with the same index value is greater than or equal to the compression threshold;

[0023] If so, the users with the same indicator value are compressed into one user group, and the indicator value is directly mapped to the user group.

[0024] Optionally, before performing rank distribution conversion on the indicator value of each user, the following steps are further included:

[0025] Determine whether the total number of users is less than the sample size threshold;

[0026] If so, the index value of each user is weighted and modified through Bayesian.

[0027] In addition, according to another aspect of an embodiment of the present invention, there is provided a device for storing user data, including:

[0028] A ranking module, used to convert the index value of each user into a rank distribution and calculate the rank rate of each user;

[0029] A mapping module, used for performing distribution mapping on the rank rate of each user based on an inverse function of a cumulative distribution function to obtain a mapping result value of each user;

[0030] An identification module, configured to determine the indicator performance level of each user based on a mapping threshold and a mapping result value of each user;

[0031] The storage module is used to store the mapping result values ​​and indicator performance levels of each user in the big data platform for the database system to call.

[0032] Optionally, the ranking module is further used to:

[0033] The index values ​​of each user are ranked with the goal of having better index value performance and smaller rank.

[0034] Optionally, the ranking module is further used to:

[0035] If the indicator value shows that the larger the indicator value, the smaller the corresponding rank, then the users are sorted in descending order of the indicator value;

[0036] If the indicator value shows that the smaller the indicator value, the smaller the corresponding rank, then the users are sorted in order from small to large indicator values.

[0037] Optionally, the ranking module is further used to: for any user, calculate the rank rate of the user by using the following method:

[0038] The rank rate of the user is calculated according to the total number of users in the ranking and the rank of the user in the ranking.

[0039] Optionally, the mapping module is further used for:

[0040] If the distribution of the index values ​​of the respective users is a skewed distribution, then distribution mapping is performed on the rank rates of the respective users based on the inverse function of the cumulative distribution function of the exponential distribution;

[0041] If the distribution of the indicator values ​​of the various users is a normal distribution, then the rank rates of the various users are distributed and mapped based on the inverse function of the cumulative distribution function of the bell-shaped distribution.

[0042] Optionally, the ranking module is further used to:

[0043] Before performing rank distribution conversion on the index value of each user, it is determined whether the number of users with the same index value is greater than or equal to the compression threshold;

[0044] If so, the users with the same indicator value are compressed into one user group, and the indicator value is directly mapped to the user group.

[0045] Optionally, the ranking module is further used to:

[0046] Before performing rank distribution transformation on the indicator values ​​of each user, determine whether the total number of users is less than the sample size threshold;

[0047] If so, the index value of each user is weighted and modified through Bayesian.

[0048] According to another aspect of an embodiment of the present invention, there is further provided an electronic device, including:

[0049] one or more processors;

[0050] a storage device for storing one or more programs,

[0051] When the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any of the above embodiments.

[0052] According to another aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored. When the program is executed by a processor, the method described in any one of the above embodiments is implemented.

[0053] One embodiment of the above invention has the following advantages or beneficial effects: because the technical means of calculating the rank ranking rate of each user and performing distribution mapping on the rank ranking rate of each user is adopted to determine the indicator performance level of each user, the technical problem of inaccurate data of each user stored in the big data platform in the prior art is overcome. The embodiment of the present invention eliminates the difference in different data dimensions in the service scenario, handles the difference in different data distributions, solves the problem of extreme value anomalies, and makes the data results very robust by performing rank distribution conversion on the indicator values; and then performs distribution mapping based on the inverse function of the cumulative distribution function to compensate for the previous information loss, so as to reduce the data information loss caused by the rank distribution conversion, thereby achieving a better data feature expression effect, thereby making the user data stored in the big data platform more accurate.

[0054] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0056] Figure 1is a schematic diagram of the main process of a method for storing user data according to an embodiment of the present invention;

[0057] Figure 2 is a schematic diagram of the main process of a method for storing user data according to a reference embodiment of the present invention;

[0058] Figure 3 is a schematic diagram of the main process of a method for storing user data according to another reference embodiment of the present invention;

[0059] Figure 4 is a schematic diagram of the main process of a method for storing user data according to yet another reference embodiment of the present invention;

[0060] Figure 5 is a schematic diagram of main modules of a device for storing user data according to an embodiment of the present invention;

[0061] Figure 6 is an exemplary system architecture diagram to which embodiments of the present invention may be applied;

[0062] Figure 7 It is a schematic diagram of the structure of a computer system of a terminal device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0064] At present, the main method used to judge the performance level of indicators is standardized. The calculation method is as follows:

[0065]

[0066] Among them, Y is the performance level of the indicator. The larger the Y is, the better the performance of the merchant in this data indicator is. X is the original value of the merchant indicator. represents the industry average level, σ is the industry standard deviation, and the calculation method is as follows:

[0067]

[0068] There are the following problems with judging the performance level of indicators through standardized methods:

[0069] 1) The method has high requirements for data distribution. It performs well only when the original data distribution is close to normal distribution and there are no extreme outliers.

[0070] 2) However, when the original data distribution is skewed (for example, the data has a serious left-leaning phenomenon), the calculated indicator performance level Y also tends to be skewed, which is often inconsistent with the actual situation. A large number of indicator performance levels Y will be lower or higher than the average level. In this case, an average merchant will have a very high performance score or a very low performance score under this indicator.

[0071] 3) When there are extreme outliers in the original data, an abnormal extreme value will affect the calculation of the industry standard deviation σ. Only one abnormal data will change the indicator performance level of all other merchants, which is not in line with the actual business situation.

[0072] 4) When only the original value of the indicator is considered, the small sample problem will lead to insufficient credibility of the merchant data indicator. For example, the data when a merchant has only made one sale is not sufficient to prove that the merchant's indicator performance level is good or bad enough.

[0073] The method for storing user data provided by the embodiment of the present invention can solve the problems of small samples, skewed indicator distribution or extreme value anomalies in data indicators, thereby obtaining a robust indicator performance level.

[0074] Figure 1 Schematic diagram of the main process of the method for storing user data according to an embodiment of the present invention. As an embodiment of the present invention, Figure 1 As shown, the method for storing user data may include:

[0075] Step 101, performing rank distribution conversion on the index value of each user, and calculating the rank rate of each user.

[0076] Different indicators have different dimensions, and their distributions vary greatly, and there are problems with extreme value outliers. Therefore, the embodiment of the present invention converts the indicator value into a rank statistic in non-parametric statistics, and then calculates the ranking rate of the rank to eliminate the dimension difference and handle the original distribution differences and outliers. For merchants on the e-commerce platform, different indicators can be configured according to the service scenario. For example, the evaluation scenario may include indicators such as product quality satisfaction, material speed satisfaction, product description satisfaction, and seller service satisfaction. The consultation scenario may include indicators such as 30-second response rate and average response time. The after-sales scenario may include indicators such as return and exchange repair rate and after-sales service time. The logistics scenario may include indicators such as 48-hour pickup timeliness rate and next-day delivery rate. The dispute scenario may include indicators such as transaction dispute rate, transaction self-service completion rate, and dispute handling compliance rate.

[0077] Optionally, the rank distribution conversion of the index value of each user includes: ranking the index value of each user with the goal of the better the index value performance, the smaller the rank ranking. That is, the better the index value X1, the smaller the rank ranking. X1 The smaller it is, the higher the rank rate X2 is.

[0078] Optionally, the index values ​​of each user are ranked with the goal of the better the index value performance and the smaller the rank ranking, including: if the index value performance is that the larger the index value, the smaller the corresponding rank ranking, then the users are ranked in descending order of the index value; if the index value performance is that the smaller the index value, the smaller the corresponding rank ranking, then the users are ranked in ascending order of the index value. This can ensure that the better the index value X1, the higher the rank ranking. X1 The smaller it is, the higher the rank rate X2 is.

[0079] Optionally, for any user, the rank rate of the user is calculated using the following method:

[0080] The rank rate of the user is calculated according to the total number of users in the ranking and the rank of the user in the ranking.

[0081] For example, the rank rate X2 can be calculated using the following formula:

[0082]

[0083] Among them, Rank X1 Indicates the rank of the user in the sorting.

[0084] At this point, the better the index value X1 performs in the industry, the greater its rank ranking rate X2. This result eliminates the dimensional differences between different data sources, is not affected by the skewness of data distribution, and has strong resistance to extreme value anomalies.

[0085] Therefore, through step 101, the original indicator value is converted into a rank statistic in non-parametric statistics, the dimension difference is eliminated, and the original distribution difference and extreme value outliers are effectively handled. Extreme value outliers refer to those that are extremely far away from the data mean and are serious outliers. They are mainly manifested as "unreasonable values" caused by special circumstances, for example, transaction dispute rate = 3000%, after-sales service time = 4000h, etc. Rank statistic is an order statistic used for non-parametric statistical tests. It is a statistic based on the size of the sample value in the total sample (rank). In actual scenarios, it can be equivalent to indicator ranking.

[0086] Step 102: Based on the inverse function of the cumulative distribution function, the rank rate of each user is distributed and mapped to obtain the mapping result value of each user.

[0087] After the rank conversion in step 101, the index difference is eliminated, but there is a problem of partial information loss. Therefore, step 102 is used to find a distribution that has a strong ability to express the original index value, and the positive and negative feedback information of the index is inferred in combination with the actual business scenario.

[0088] After determining the target distribution, since the rank ranking rate X2 after rank transformation is approximately uniformly distributed, the inverse function mapping method of the cumulative distribution function F of the target distribution can be used to transform the distribution of the rank ranking rate X2 into the target distribution. The specific formula is as follows:

[0089] X3=F -1 (X2), F -1 Represents the inverse function of F

[0090] Inverse function distribution mapping: If the cumulative distribution function F is a continuous strictly increasing function, then there exists its inverse function F -1 (y), y∈[0,1]. The inverse function of the cumulative distribution function can be used to generate random variables that obey this random distribution. Assume that F X (x) is the cumulative distribution function of the probability distribution X, and there exists an inverse function If a is a random variable uniformly distributed in the interval [0,1), then It follows the distribution of X.

[0091] Afterwards, the threshold of the mapping result value X3 is limited according to the actual business scenario, and the mapping result value X3 is observed to see whether it satisfies the positive and negative feedback of the business for the indicator in the business scenario. If it does, the mapping is performed with the target distribution. If it does not, the target distribution is searched again until the mapping result value X3 satisfies the positive and negative feedback of the business for the indicator in the business scenario. Positive and negative feedback refers to feedback information obtained based on input information. In an embodiment of the present invention, it refers to the result value obtained after the original indicator value of the indicator is transformed by rank distribution and mapped by inverse function distribution. When the result value is greater than 0, it indicates that the indicator is positive in the scenario. The larger the result value, the better the indicator performs in the scenario, and vice versa.

[0092] Optionally, if the distribution of the index values ​​of the respective users is a skewed distribution, then the rank ranking rate of the respective users is distributed and mapped based on the inverse function of the cumulative distribution function of the exponential distribution; if the distribution of the index values ​​of the respective users is a normal distribution, then the rank ranking rate of the respective users is distributed and mapped based on the inverse function of the cumulative distribution function of the bell-shaped distribution. Skewed distribution: In contrast to the "normal distribution", the data frequency distribution with an asymmetric distribution curve is a type of probability distribution of continuous random variables. The degree of skewness can be measured by calculating the kurtosis and skewness.

[0093] Generally, for the original indicator value distribution that is severely skewed (such as transaction dispute rate), an exponential distribution mapping can be used, as shown below:

[0094] X3=0.7*ln(X2)

[0095] Generally, for original indicator values ​​with relatively symmetrical distribution (such as next-day delivery rate), a bell-shaped distribution mapping can be used, as shown below:

[0096]

[0097] In an embodiment of the present invention, the process of transforming from rank distribution to inverse function distribution mapping is collectively referred to as positive and negative feedback, which means judging whether the indicator is positive or negative feedback for the user based on the original value of the user indicator, and judging the feedback intensity.

[0098] Step 103: determining the indicator performance level of each user based on the mapping threshold and the mapping result value of each user.

[0099] Generally speaking, the size of the mapping threshold can be set according to the target distribution and the confidence interval. Taking the confidence interval of 95% as an example, if the target distribution is an exponential distribution, the mapping threshold is [-2,0], and if the target distribution is a bell-shaped distribution, the mapping threshold is [-2,2]. After obtaining the mapping result value, the indicator performance level of each user is analyzed based on the corresponding mapping threshold.

[0100] Optionally, the recording results of the performance level can be divided into two types: one type is the feedback value, which uses a numerical value to express the intensity of positive and negative feedback. The more it is greater than 0, the better the performance of the indicator, and vice versa; the other type is the feedback type, which uses a classification result to express the judgment of the indicator, such as excellent, good, medium, poor, and inferior.

[0101] Step 104, storing the mapping result values ​​and indicator performance levels of each user in the big data platform for the database system to call.

[0102] After obtaining the mapping result value and indicator performance level of each user, the mapping result value and indicator performance level can be stored in the big data platform for direct call by the database system to meet the feature extraction requirements of other models.

[0103] According to the various embodiments described above, it can be seen that the present invention solves the technical problem of inaccurate data of each user stored in the big data platform in the prior art by calculating the rank ranking rate of each user and performing distribution mapping on the rank ranking rate of each user, thereby determining the indicator performance level of each user. The embodiment of the present invention eliminates the differences in different data dimensions in the service scenario, handles the differences in different data distributions, solves the problem of extreme value anomalies, and makes the data results very robust; and then compensates for the previous information loss by performing distribution mapping based on the inverse function of the cumulative distribution function, so as to reduce the data information loss caused by the rank distribution conversion, thereby achieving a better data feature expression effect, thereby making the user data stored in the big data platform more accurate.

[0104] Figure 2 FIG. 1 is a schematic diagram of the main process of a method for storing user data according to a reference embodiment of the present invention. As another embodiment of the present invention, the method for storing user data may include the following steps:

[0105] Step 201 , determining whether there are users whose number of users with the same index value is greater than or equal to a compression threshold; if so, executing step 202 ; if not, executing step 203 .

[0106] Step 202: compress the users with the same indicator value into a user group, and directly map the indicator value to the user group.

[0107] Before performing rank distribution conversion, we first consider the node problem of rank distribution, that is, whether there are a large number of users with the same index value. If such special nodes exist, we use node compression to directly map the same index value to a suitable rank. X1 That is, users with the same index value are compressed into a user group, and the index value is directly mapped to the user group. When the index value is rank-distributed, the user group is ranked as a user, then the rank of each user in the user group is X1 The same, the rank rate X2 is also the same.

[0108] Step 203, with the goal of having a better index value and a smaller rank, the index values ​​of each user are ranked.

[0109] Optionally, the index values ​​of each user are ranked with the goal of the better the index value performance and the smaller the rank ranking, including: if the index value performance is that the larger the index value, the smaller the corresponding rank ranking, then the users are ranked in descending order of the index value; if the index value performance is that the smaller the index value, the smaller the corresponding rank ranking, then the users are ranked in ascending order of the index value. This can ensure that the better the index value X1, the higher the rank ranking. X1 The smaller it is, the higher the rank rate X2 is.

[0110] Step 204: Calculate the rank rate of each user.

[0111] The rank rate X2 can be calculated using the following formula:

[0112]

[0113] Among them, Rank X1 Indicates the rank of the user in the sorting.

[0114] Step 205 : Based on the inverse function of the cumulative distribution function, the rank rate of each user is distributed and mapped to obtain the mapping result value of each user.

[0115] If the distribution of the indicator values ​​of each user is a skewed distribution, the rank ranking rate of each user is distributed and mapped based on the inverse function of the cumulative distribution function of the exponential distribution; if the distribution of the indicator values ​​of each user is a normal distribution, the rank ranking rate of each user is distributed and mapped based on the inverse function of the cumulative distribution function of the bell-shaped distribution.

[0116] Step 206: Determine the indicator performance level of each user based on the mapping threshold and the mapping result value of each user.

[0117] Step 207, storing the mapping result value and indicator performance level of each user in the big data platform for the database system to call.

[0118] In addition, the specific implementation content of the method for storing user data in a reference embodiment of the present invention has been described in detail in the method for storing user data described above, so the repeated content will not be described here.

[0119] Figure 3 1 is a schematic diagram of the main process of a method for storing user data according to another reference embodiment of the present invention. As another embodiment of the present invention, the method for storing user data may include the following steps:

[0120] Step 301, determine whether the total number of users is less than the sample size threshold; if so, execute step 302; if not, execute step 303.

[0121] Step 302: weighted correction is performed on the index value of each user through Bayesian method.

[0122] If the sample size of the indicator is low (that is, the total number of users is small, generally speaking, there are as many samples as there are users), the credibility of the indicator result provided is also low, and the result value fluctuates greatly, which is not conducive to identifying the indicator performance level. Therefore, in step 302, according to the Bayesian theory, the results of the small sample data are corrected to improve the overall stability of the model.

[0123] Optionally, the Bayesian weighted correction formula is as follows:

[0124]

[0125] m c =n t -n0

[0126] Among them, X0 represents the index value before correction, n0 represents the sample size of the index; P represents the average index value of the industry; parameter n t Determined by the actual business scenario, it represents the minimum sample size for reliable indicator results. The default value can be set to 15, 20, 30, 40 or 50, etc.

[0127] Bayes' theorem is a theorem in probability theory that describes the probability of an event under certain known conditions. Bayesian weighted correction refers to adjusting the indicator results based on actual sample data while assuming that the indicator results are the industry average. The embodiment of the present invention corrects small sample data based on Bayesian weighting to solve the problem of low credibility and insufficient business value of small sample data.

[0128] Step 303, ranking the index values ​​of each user with the goal of having better index values ​​and smaller rank.

[0129] Optionally, the index values ​​of each user are ranked with the goal of the better the index value performance and the smaller the rank ranking, including: if the index value performance is that the larger the index value, the smaller the corresponding rank ranking, then the users are ranked in descending order of the index value; if the index value performance is that the smaller the index value, the smaller the corresponding rank ranking, then the users are ranked in ascending order of the index value. This can ensure that the better the index value X1, the higher the rank ranking. X1 The smaller it is, the higher the rank rate X2 is.

[0130] Step 304: Calculate the rank rate of each user.

[0131] The rank rate X2 can be calculated using the following formula:

[0132]

[0133] Among them, Rank X1 Indicates the rank of the user in the sorting.

[0134] Step 305: Based on the inverse function of the cumulative distribution function, the rank rate of each user is distributed and mapped to obtain the mapping result value of each user.

[0135] For the original indicator values ​​with serious skewness (such as transaction dispute rate), exponential distribution mapping can be used, as shown below:

[0136] X3=0.7*ln(X2)

[0137] For original indicator values ​​with relatively symmetrical distribution (such as next-day delivery rate), a bell-shaped distribution mapping can be used, as shown below:

[0138]

[0139] Step 306: Determine the indicator performance level of each user based on the mapping threshold and the mapping result value of each user.

[0140] Step 307: storing the mapping result values ​​and indicator performance levels of each user in the big data platform for the database system to call.

[0141] In addition, the specific implementation content of the method for storing user data in another reference embodiment of the present invention has been described in detail in the method for storing user data described above, so the repeated content will not be described here.

[0142] Figure 4 1 is a schematic diagram of the main process of a method for storing user data according to another reference embodiment of the present invention. As another embodiment of the present invention, the method for storing user data may include the following steps:

[0143] Step 401, calculate the industry mean, variance, and extreme value of the indicator item.

[0144] Step 402, determine whether there are significant differences in the industry mean, variance, and extreme value of the indicator items; if not, end; if so, execute step 403.

[0145] Only when the selected indicator has information difference, the indicator can provide effective information for subsequent use. Therefore, in step 402, it is determined whether the industry information of the indicator is sufficient and logical fallacies are eliminated to improve the overall performance of the model.

[0146] In the embodiment of the present invention, it is mainly considered whether there are significant differences in the industry mean, variance, and extreme value of the indicator items, and the judgment logic is as follows:

[0147] Industry average

[0148] Industry variance

[0149] Industry Extreme Value X t =max(X)-min(X)<threshold C

[0150] The thresholds A, B, and C are obtained based on the actual data distribution evaluation. For example, the default values ​​of A, B, and C can be 0, 0.0001, and 0.01.

[0151] If the above three conditions are met at the same time, it means whether there are significant differences in the industry mean, variance and extreme value of the indicator item.

[0152] Step 403, with the goal of having a better index value and a smaller rank, the index values ​​of each user are ranked.

[0153] Step 404: Calculate the rank rate of each user.

[0154] Step 405: Based on the inverse function of the cumulative distribution function, the rank rate of each user is distributed and mapped to obtain the mapping result value of each user.

[0155] Step 406: Determine the indicator performance level of each user based on the mapping threshold and the mapping result value of each user.

[0156] Step 407, storing the mapping result value and indicator performance level of each user in the big data platform for the database system to call.

[0157] In addition, the specific implementation content of the method for storing user data in another reference embodiment of the present invention has been described in detail in the method for storing user data described above, so the repeated content will not be described here.

[0158] Figure 5 is a schematic diagram of main modules of a device for storing user data according to an embodiment of the present invention, such as Figure 5As shown, the device 500 for storing user data includes a ranking module 501, a mapping module 502, an identification module 503 and a storage module 504. Among them, the ranking module 501 is used to perform rank distribution conversion on the index value of each user and calculate the rank ranking rate of each user; the mapping module 502 is used to perform distribution mapping on the rank ranking rate of each user based on the inverse function of the cumulative distribution function to obtain the mapping result value of each user; the identification module 503 is used to determine the index performance level of each user based on the mapping threshold and the mapping result value of each user; the storage module 504 is used to store the mapping result value and the index performance level of each user in the big data platform for the database system to call.

[0159] Optionally, the ranking module 501 is further used for:

[0160] The index values ​​of each user are ranked with the goal of having better index value performance and smaller rank.

[0161] Optionally, the ranking module 501 is further used for:

[0162] If the indicator value shows that the larger the indicator value, the smaller the corresponding rank, then the users are sorted in descending order of the indicator value;

[0163] If the indicator value shows that the smaller the indicator value, the smaller the corresponding rank, then the users are sorted in order from small to large indicator values.

[0164] Optionally, the ranking module 501 is further used to: for any user, calculate the rank rate of the user by using the following method:

[0165] The rank rate of the user is calculated according to the total number of users in the ranking and the rank of the user in the ranking.

[0166] Optionally, the mapping module 502 is further configured to:

[0167] If the distribution of the index values ​​of the respective users is a skewed distribution, a distribution mapping is performed on the rank rates of the respective users based on an inverse function of the cumulative distribution function of the exponential distribution;

[0168] If the distribution of the indicator values ​​of the various users is a normal distribution, then the rank rates of the various users are distributed and mapped based on the inverse function of the cumulative distribution function of the bell-shaped distribution.

[0169] Optionally, the ranking module 501 is further used for:

[0170] Before performing rank distribution conversion on the index value of each user, it is determined whether the number of users with the same index value is greater than or equal to the compression threshold;

[0171] If so, the users with the same indicator value are compressed into one user group, and the indicator value is directly mapped to the user group.

[0172] Optionally, the ranking module 501 is further used for:

[0173] Before performing rank distribution transformation on the indicator values ​​of each user, determine whether the total number of users is less than the sample size threshold;

[0174] If so, the index value of each user is weighted and modified through Bayesian.

[0175] According to the various embodiments described above, it can be seen that the present invention solves the technical problem of inaccurate data of each user stored in the big data platform in the prior art by calculating the rank ranking rate of each user and performing distribution mapping on the rank ranking rate of each user, thereby determining the indicator performance level of each user. The embodiment of the present invention eliminates the differences in different data dimensions in the service scenario, handles the differences in different data distributions, solves the problem of extreme value anomalies, and makes the data results very robust; and then compensates for the previous information loss by performing distribution mapping based on the inverse function of the cumulative distribution function, so as to reduce the data information loss caused by the rank distribution conversion, thereby achieving a better data feature expression effect, thereby making the user data stored in the big data platform more accurate.

[0176] It should be noted that the specific implementation content of the device for storing user data in the present invention has been described in detail in the method for storing user data described above, so the repeated content will not be described again here.

[0177] Figure 6 An exemplary system architecture 600 to which the method for storing user data or the apparatus for storing user data according to the embodiment of the present invention can be applied is shown.

[0178] like Figure 6 As shown, system architecture 600 may include terminal devices 601, 602, 603, network 604 and server 605. Network 604 is used to provide a medium for communication links between terminal devices 601, 602, 603 and server 605. Network 604 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0179] Users can use terminal devices 601, 602, and 603 to interact with server 605 through network 604 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 601, 602, and 603, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only examples).

[0180] The terminal devices 601 , 602 , and 603 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers, etc.

[0181] The server 605 may be a server that provides various services, such as a backend management server (only an example) that provides support for shopping websites browsed by users using the terminal devices 601, 602, and 603. The backend management server may analyze and process the received data such as the item information query request, and feed back the processing results (such as target push information, item information - only an example) to the terminal device.

[0182] It should be noted that the method for storing user data provided in the embodiment of the present invention is generally performed by the server 605, and accordingly, the device for storing user data is generally set in the server 605. The method for storing user data provided in the embodiment of the present invention can also be performed by the terminal devices 601, 602, and 603, and accordingly, the device for storing user data can be set in the terminal devices 601, 602, and 603.

[0183] It should be understood that Figure 6 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0184] Reference below Figure 7 , which shows a schematic diagram of the structure of a computer system 700 of a terminal device suitable for implementing an embodiment of the present invention. Figure 7 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0185] like Figure 7As shown, the computer system 700 includes a central processing unit (CPU) 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage part 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the system 700 are also stored. The CPU 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0186] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, a mouse, etc.; an output section 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, a modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as needed, so that a computer program read therefrom is installed into the storage section 708 as needed.

[0187] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 709, and / or installed from the removable medium 711. When the computer program is executed by the central processing unit (CPU) 701, the above-mentioned functions defined in the system of the present invention are executed.

[0188] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0189] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0190] The modules involved in the embodiments of the present invention may be implemented by software or hardware. The modules described may also be set in a processor, for example, it may be described as: a processor includes a ranking module, a mapping module and an identification module, wherein the names of these modules do not constitute a limitation on the modules themselves in some cases.

[0191] As another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiment; or it may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device includes: performing rank distribution conversion on the indicator value of each user, and calculating the rank ranking rate of each user; based on the inverse function of the cumulative distribution function, performing distribution mapping on the rank ranking rate of each user, and obtaining the mapping result value of each user; based on the mapping threshold and the mapping result value of each user, determining the indicator performance level of each user; storing the mapping result value and indicator performance level of each user in the big data platform for the database system to call.

[0192] According to the technical solution of the embodiment of the present invention, because the technical means of calculating the rank ranking rate of each user and performing distribution mapping on the rank ranking rate of each user is adopted to determine the indicator performance level of each user, the technical problem of inaccurate data of each user stored in the big data platform in the prior art is overcome. The embodiment of the present invention eliminates the difference in different data dimensions in the service scenario, handles the difference in different data distributions, solves the problem of extreme value anomalies, and makes the data results very robust by performing rank distribution conversion on the indicator values; and then compensates for the previous information loss by performing distribution mapping based on the inverse function of the cumulative distribution function to reduce the data information loss caused by the rank distribution conversion, thereby achieving a better data feature expression effect, thereby making the user data stored in the big data platform more accurate.

[0193] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for storing user data, characterized in that: include: Perform rank distribution conversion on the indicator value of each user, and calculate the rank rate of each user; Based on the inverse function of the cumulative distribution function, the rank rate of each user is distributed and mapped to obtain the mapping result value of each user; Determining the indicator performance level of each user based on the mapping threshold and the mapping result value of each user; The mapping result value and the indicator performance level of each user are stored in the big data platform for the database system to call; the performance level is a feedback value or a feedback type; Based on the inverse function of the cumulative distribution function, the rank rate of each user is distributed and mapped, including: If the distribution of the index values ​​of the respective users is a skewed distribution, a distribution mapping is performed on the rank rates of the respective users based on an inverse function of the cumulative distribution function of the exponential distribution; If the distribution of the indicator values ​​of the respective users is a normal distribution, then distribution mapping is performed on the rank rates of the respective users based on the inverse function of the cumulative distribution function of the bell-shaped distribution; In the evaluation scenario, the indicators include satisfaction with product quality, satisfaction with material speed, satisfaction with product description, and satisfaction with seller service; In the case of consultation scenarios, the indicators include 30-second response rate and average response time; In the after-sales scenario, the indicators include return and exchange repair rate and after-sales service time; In the logistics scenario, the indicators include the 48-hour pickup rate and the next-day delivery rate; In the case of a dispute scenario, the indicators include transaction dispute rate, transaction self-service completion rate, and dispute resolution compliance rate.

2. The method according to claim 1, characterized in that Perform rank distribution transformation on the indicator values ​​of each user, including: The index values ​​of each user are ranked with the goal of having better index value performance and smaller rank.

3. The method according to claim 2, characterized in that With the goal of better index value performance and smaller rank, the index values ​​of each user are ranked, including: If the indicator value shows that the larger the indicator value, the smaller the corresponding rank, then the users are sorted in descending order of the indicator value; If the indicator value shows that the smaller the indicator value, the smaller the corresponding rank, then the users are sorted in order from small to large indicator values.

4. The method according to claim 1, characterized in that For any user, the rank rate of the user is calculated using the following method: The rank rate of the user is calculated according to the total number of users in the ranking and the rank of the user in the ranking.

5. The method according to claim 1, characterized in that Before the rank distribution transformation of the indicator value of each user is performed, it also includes: Determine whether there are users whose number of users with the same index value is greater than or equal to the compression threshold; If so, the users with the same indicator value are compressed into one user group, and the indicator value is directly mapped to the user group.

6. The method according to claim 1, characterized in that Before the rank distribution transformation of the indicator value of each user is performed, it also includes: Determine whether the total number of users is less than the sample size threshold; If so, the index value of each user is weighted and modified through Bayesian.

7. A device for storing user data, characterized in that: include: A ranking module, used to convert the index value of each user into a rank distribution and calculate the rank rate of each user; A mapping module, used for performing distribution mapping on the rank rate of each user based on an inverse function of a cumulative distribution function to obtain a mapping result value of each user; an identification module, configured to determine the indicator performance level of each user based on a mapping threshold and a mapping result value of each user; the performance level is a feedback value or a feedback type; A storage module, used to store the mapping result values ​​and indicator performance levels of each user in the big data platform for the database system to call; The mapping module is also used to: If the distribution of the index values ​​of the respective users is a skewed distribution, a distribution mapping is performed on the rank rates of the respective users based on an inverse function of the cumulative distribution function of the exponential distribution; If the distribution of the indicator values ​​of the respective users is a normal distribution, then distribution mapping is performed on the rank rates of the respective users based on the inverse function of the cumulative distribution function of the bell-shaped distribution; In the evaluation scenario, the indicators include satisfaction with product quality, satisfaction with material speed, satisfaction with product description, and satisfaction with seller service; In the case of consultation scenarios, the indicators include 30-second response rate and average response time; In the after-sales scenario, the indicators include return and exchange repair rate and after-sales service time; In the logistics scenario, the indicators include the 48-hour pickup rate and the next-day delivery rate; In the case of a dispute scenario, the indicators include transaction dispute rate, transaction self-service completion rate, and dispute resolution compliance rate.

8. An electronic device, characterized in that: include: one or more processors; a 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 method according to any one of claims 1 to 6.

9. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images