User loyalty scoring method, device, equipment and readable storage medium

Abstract

The embodiment of the present application provides a user loyalty scoring method, device, electronic device and readable storage medium, which relates to the field of data processing. The method includes: obtaining multiple sample data; presetting each parameter in the multiple sample data to obtain an index for evaluating user loyalty; for each sample data, using the entropy method to determine the proportion of each index in the sample data in the multiple sample data; for each user, based on the proportion and weight of each index of the sample data corresponding to the user, determining the user's first loyalty score; using the factor analysis method to determine the correlation between each index in each multiple sample data, determine each scoring factor and the score of each scoring factor corresponding to each sample data; for each user, based on the score of each scoring factor corresponding to the user, determine the user's second loyalty score, and based on the user's first loyalty score and the second loyalty score, determine the user's target score.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a user loyalty scoring method, device, electronic device and readable storage medium. Background Art

[0002] Loyal users can not only create value for the enterprise, but also bring a lot of potential benefits to the enterprise, such as brand promotion, driving other customers to buy and grow, etc. Therefore, it is necessary to analyze and evaluate user loyalty and then make corresponding marketing strategies. With the goal of improving customer satisfaction and expanding the customer base, it helps enterprises achieve continuous growth in marketing and profits.

[0003] In the current user loyalty scoring scheme, the scoring algorithm used is single, resulting in low accuracy of the scoring results. Summary of the invention

[0004] The embodiment of the present application provides a user loyalty scoring method, device, electronic device and readable storage medium, which can solve the problem of low accuracy of user loyalty scoring. The technical solution is as follows:

[0005] According to one aspect of an embodiment of the present application, a method for scoring user loyalty is provided, the method comprising:

[0006] Acquire multiple sample data, each sample data corresponds to a user, and each sample data includes multiple parameters;

[0007] Performing preset processing on each parameter in the plurality of sample data to obtain an index for evaluating user loyalty; wherein the preset processing includes: parameter screening and parameter derivation;

[0008] For each sample data, an entropy method is used to determine the proportion of each indicator in the sample data in the multiple sample data;

[0009] For each user, based on the proportion of each indicator of the sample data corresponding to the user and the weight of each indicator, determine a first loyalty score of the user;

[0010] Using factor analysis to determine the correlation between the indicators in the plurality of sample data, and to determine the scoring factors and the scores of the scoring factors corresponding to the sample data;

[0011] For each user, determining a second loyalty score of the user based on the scores of the scoring factors corresponding to the user;

[0012] For each user, a target score of the user is determined based on the first loyalty score and the second loyalty score of the user.

[0013] In a possible implementation, before the step of using factor analysis to determine the correlation between the indicators in the plurality of sample data and determining the scoring factors and the scores of the scoring factors corresponding to the sample data, the method further includes:

[0014] Performing standardization processing on the plurality of sample data to obtain a data set;

[0015] The negative data in the data set are subjected to translation processing to obtain non-negative data, wherein the translation amount during the translation processing is the sum of a preset value and the absolute value of the minimum data in the data set.

[0016] In another possible implementation, before the step of determining, for each user, the target score of the user based on the first loyalty score and the second loyalty score of the user, the method further includes:

[0017] Performing consistency check on the first loyalty scores and the second loyalty scores of all users corresponding to the plurality of sample data;

[0018] In the case where the consistency check passes, the step of determining, for each user, a target score of the user based on the first loyalty score and the second loyalty score of the user is performed.

[0019] In yet another possible implementation, the step of performing consistency verification on the first loyalty scores and the second loyalty scores of all users corresponding to the plurality of sample data includes:

[0020] Based on the first loyalty score and the second loyalty score of all users, the intraclass correlation coefficient ICC is calculated;

[0021] If the intraclass correlation coefficient ICC is within a preset range, it is determined that the first loyalty scores and the second loyalty scores of all users pass the consistency check.

[0022] In yet another possible implementation, the step of determining, for each user, the second loyalty score of the user based on the scores of the scoring factors corresponding to the user, includes:

[0023] For each user, the variance contribution rate of each scoring factor corresponding to the user is used as the weight of each scoring factor;

[0024] The scores of the scoring factors are weighted and calculated to obtain the second loyalty score of the user.

[0025] In yet another possible implementation, before the step of determining, for each sample data, the proportion of each indicator in the sample data in the plurality of sample data using an entropy method, the step further includes:

[0026] Determine the labeling information of multiple indicators corresponding to each sample data according to business experience to determine positive indicators and negative indicators, wherein the positive indicators have a positive impact on the user's loyalty score, and the negative indicators have a negative impact on the user's loyalty score;

[0027] Bind the sample data corresponding to each user and the annotation information of multiple indicators to obtain the characteristic data of the corresponding user;

[0028] Preprocessing is performed on the characteristic data of each user, and the preprocessing includes: using different normalization methods to process positive indicators and negative indicators;

[0029] The step of using the entropy method to determine the proportion of each indicator in the sample data in the plurality of sample data for each sample data includes:

[0030] For each user's characteristic data, an entropy method is used to determine the proportion of each indicator in the characteristic data in the characteristic data of all users.

[0031] According to another aspect of an embodiment of the present application, a user loyalty scoring device is provided, the device comprising:

[0032] The acquisition module is used to obtain multiple sample data, each sample data corresponds to a user

[0033] , each sample data includes multiple parameters;

[0034] A preprocessing module, used to perform preset processing on each parameter in the plurality of sample data to obtain an index for evaluating user loyalty; wherein the preset processing includes: parameter screening and parameter derivation;

[0035] A determination module is used to determine, for each sample data, the proportion of each indicator in the sample data in the multiple sample data by using an entropy method; it is also used to determine, for each user, the first loyalty score of the user based on the proportion of each indicator in the sample data corresponding to the user and the weight of each indicator; it is also used to determine, by using a factor analysis method, the correlation between each indicator in the multiple sample data, determine each scoring factor and the score of each scoring factor corresponding to each sample data; it is also used to determine, for each user, the second loyalty score of the user based on the score of each scoring factor corresponding to the user; for each user, determine the target score of the user based on the first loyalty score and the second loyalty score of the user.

[0036] In a possible implementation, the preprocessing module is further used to:

[0037] Performing standardization processing on the plurality of sample data to obtain a data set;

[0038] The negative data in the data set are subjected to translation processing to obtain non-negative data, wherein the translation amount during the translation processing is the sum of a preset value and the absolute value of the minimum data in the data set.

[0039] According to another aspect of an embodiment of the present application, an electronic device is provided. The electronic device includes a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the steps of the above-mentioned user loyalty scoring method.

[0040] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned user loyalty scoring method are implemented.

[0041] The beneficial effects of the technical solution provided by the embodiment of the present application are:

[0042] In the embodiment of the present application, the user loyalty score is scored by using factor analysis and entropy method respectively, and the user loyalty score is obtained based on the two scores, thereby solving the problem of low accuracy of user loyalty score. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in describing the embodiments of the present application are briefly introduced below.

[0044] Figure 1 A flowchart of a user loyalty scoring method provided in an embodiment of the present application;

[0045] Figure 2 A schematic diagram of the structure of a user loyalty scoring device provided in an embodiment of the present application;

[0046] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0047] The embodiments of the present application are described below in conjunction with the drawings in the present application. It should be understood that the implementation methods described below in conjunction with the drawings are exemplary descriptions for explaining the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions of the embodiments of the present application.

[0048] It will be understood by those skilled in the art that, unless specifically stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the terms "including" and "comprising" used in the embodiments of the present application refer to that the corresponding features can be implemented as the presented features, information, data, steps, operations, elements and / or components, but do not exclude the implementation as other features, information, data, steps, operations, elements, components and / or combinations thereof supported by the technical field. It should be understood that when we say that an element is "connected" or "coupled" to another element, the one element may be directly connected or coupled to the other element, or it may refer to that the one element and the other element establish a connection relationship through an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein indicates at least one of the items defined by the term, for example, "A and / or B" may be implemented as "A", or as "B", or as "A and B".

[0049] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0050] At present, there are two main methods for user loyalty scoring: subjective weighting method and objective weighting method. Among them, subjective weighting method includes hierarchical analysis method, expert survey method, etc., and objective weighting method includes principal component analysis, factor analysis, entropy method, etc. Since most of the existing scoring schemes only use a single method for scoring, there may be bias, resulting in inaccurate scoring.

[0051] In information theory, entropy is a measure of uncertainty. Entropy can be used to determine the degree of dispersion of an indicator. The greater the degree of dispersion of an indicator, the greater the impact of the indicator on the comprehensive evaluation. The entropy method is a method that uses the information entropy of the indicator to determine the weight coefficient of each indicator. However, the entropy method is a model that determines the weight by the degree of dispersion of a single indicator, and lacks horizontal comparison between indicators.

[0052] Factor analysis is a method of grouping original variables according to the size of correlation, so that the correlation between variables in the same group is higher, while the correlation between variables in different groups is lower. The variables with complex relationships between indicators are reduced to a few comprehensive factors, and the contribution of each factor is used as a weight to calculate the score of each sample.

[0053] When aggregating two models, we need to consider whether the data of the two models are suitable for aggregation (whether there is high intra-group consistency). The basic function of the intra-group correlation coefficient (ICC) is to see whether the individual data can be properly aggregated into higher-order data for analysis at a higher order.

[0054] When using the entropy method to score users, it was found that there were large deviations in the results. Considering that the entropy method determines the weights based on the vertical discreteness of the indicators and lacks horizontal comparison between indicators, the factor analysis method is added on the basis of the entropy method. The factor analysis method uses the correlation between indicators to calculate the factor score, which solves the defects of the entropy method.

[0055] Based on the above findings, the present application provides a method for scoring user loyalty, which uses the improved factor analysis method and the entropy method to score user loyalty respectively, and then uses the ICC intra-group correlation coefficient to test whether the two scoring results are consistent. Finally, the two scoring results are aggregated to obtain the user's loyalty score, thereby solving the problem of low accuracy of user loyalty scoring.

[0056] In addition, by comparing the scoring results of different algorithms and the customer churn, it is found that the scoring results aggregated using the improved factor analysis method and the entropy method are more objective and fair. Therefore, the solution provided in this application is more adaptable.

[0057] The following describes several exemplary embodiments to illustrate the technical solutions of the embodiments of the present application and the technical effects produced by the technical solutions of the present application. It should be noted that the following embodiments can refer to, draw on or combine with each other, and the same terms, similar features and similar implementation steps in different embodiments will not be described repeatedly.

[0058] The present application provides a user loyalty scoring method, which can be applied to various scenarios where user loyalty needs to be evaluated, such as banks, e-commerce platforms, operators, etc. Figure 1 As shown, the method includes:

[0059] S101. Acquire a plurality of sample data, where each sample data in the plurality of sample data corresponds to a user, and each sample data includes a plurality of parameters.

[0060] Specifically, in this embodiment, each sample data includes: user basic data, user behavior data and business data.

[0061] If the user is a bank depositor, the basic data may include: account opening time, age, gender, user level (gold card user, silver card user, VIP card, etc.); user behavior data may include: consumption (swiping card, payment, etc.), transfer (transfer out, transfer in, etc.); business data may include: which business products have been handled (wealth management, fixed deposits, etc.), how many of each business product have been handled, etc.

[0062] If the user is a mobile / telecom user, the basic data may include: network age, age, gender, user level (5-star user, 3-star user, 2-star user, etc.); user behavior data may include: recharging phone bills, recharging traffic, etc.; business data may include: which business products have been applied for (traffic packages, daily packages, monthly packages, half-year packages, annual packages, etc.), how many of each business product have been applied for, etc.

[0063] S102, performing preset processing on each parameter in the plurality of sample data to obtain an index for evaluating user loyalty; wherein the preset processing includes: parameter screening and parameter derivation.

[0064] Specifically, in this embodiment, the indicators for evaluating user loyalty may include multiple first indicators for evaluating user loyalty directly selected from sample data and at least one second indicator (month-on-month, year-on-year, average, etc.) derived from the multiple first indicators. If the user is a bank depositor, the first indicator may include: deposits, loans, financial management, transaction amount, etc.

[0065] S103: For each sample data, use an entropy method to determine the proportion of each indicator in the sample data in the plurality of sample data.

[0066] S104: For each user, determine a first loyalty score of the user based on the proportion of each indicator of the sample data corresponding to the user and the weight of each indicator.

[0067] Specifically, in this embodiment, assuming that there are N users and M indicators, the following takes the jth indicator of the ith user as an example to describe the process of calculating the first loyalty score of the ith user by using the entropy method.

[0068] First, calculate the proportion Pij of the value x of the jth index of the i-th user to the value of the jth index of all users. Specifically, it can be calculated by the following formula (1):

[0069] Pij=Xij / sum(Xij) (1)

[0070] Among them, sum(Xij) is the value of the j-th indicator of all users, and when summing, the value of i ranges from 1 to N.

[0071] It should be understood that when i takes a value between 1 and N, and j takes a value between 1 and M, the proportion of the M indicators of the i-th user can be calculated; when i takes a value between 1 and N, and j takes a value between 1 and M, the proportion of the M indicators of each user among the N users can be calculated.

[0072] Secondly, the weight Wj of the jth indicator is calculated based on the weight Pij. Specifically, it can be calculated by the following formulas (2)-(4):

[0073] ej=-1 / ln(M)*sum(Pij*ln(Pij)) (2)

[0074] dj=1-ej (3)

[0075] Wj=dj / sum(dj) (4)

[0076] Among them, ej is the entropy value of the jth indicator, the user quantity is M, ln(Pij) is the logarithm of the weight Pij, and dj is the information entropy redundancy of the jth indicator. Information entropy redundancy refers to the repetition of information, which directly affects the size of the weight. The greater the information entropy redundancy, the greater the importance to the evaluation and the greater the weight.

[0077] When i takes a value from 1 to N, and j takes a value from 1 to M, the entropy values ​​of the M indicators of the i-th user can be calculated, and then the information entropy redundancy and weight of the M indicators can be obtained.

[0078] Finally, based on the proportions and weights of all indicators of the i-th user, the first loyalty score fi of the i-th user is calculated. Specifically, it can be calculated by the following formula (5):

[0079] fi=sum(Wj*Pij) (5)

[0080] When i takes a value from 1 to N and j takes a value from 1 to M, the product of the weight and proportion of each indicator of the i-th user can be calculated, and then the sum of the products of the weights and proportions of the M indicators can be calculated to obtain the first loyalty score fi of the i-th user.

[0081] It should be understood that when i takes any value from 1 to N, the above formulas (1) to (5) can be used to calculate the first loyalty score fi of each user among the N users.

[0082] S105: Using factor analysis method, determine the correlation between the indicators in the plurality of sample data, and determine the scoring factors and the scores of the scoring factors corresponding to the sample data.

[0083] Specifically, in this embodiment, the correlation coefficient matrix of each indicator in the sample data can be calculated, the factor loading matrix can be calculated, the common factors for scoring can be selected, the factor rotation can be performed, and the scores of the single common factors can be calculated to obtain the scores of each scoring factor.

[0084] S106: For each user, determine a second loyalty score of the user based on the scores of the scoring factors corresponding to the user.

[0085] Specifically, in this embodiment, the variance contribution rate of each common factor can be used as a weight to weight the common factor score, and the comprehensive factor score corresponding to each sample data is calculated, that is, the second loyalty score of each user.

[0086] S107. For each user, determine a target score of the user based on the first loyalty score and the second loyalty score of the user.

[0087] That is to say, in the embodiment of the present application, the user loyalty score is scored by using the factor analysis method and the entropy method respectively, and the user loyalty score is obtained based on the two scores, thereby solving the problem of low accuracy of the user loyalty score.

[0088] A possible implementation method is also provided in the embodiment of the present application. Step S106 may specifically include:

[0089] S1061. For each user, the variance contribution rate of each scoring factor corresponding to the user is used as the weight of each scoring factor.

[0090] S1062: Perform weighted calculation on the scores of the scoring factors to obtain a second loyalty score of the user.

[0091] Specifically, in this embodiment, assuming that the i-th user has K scoring factors, the variance contribution rates of the K scoring factors can be used as the weights of the K scoring factors. If the variance contribution rate of the q-th scoring factor is the weight Wq of the q-th scoring factor, and the value of q ranges from 1 to K, then the second loyalty score Fi of the i-th user is Fi=sum(Wq*Vq), where Vq is the score of the q-th scoring factor.

[0092] A possible implementation method is also provided in the embodiment of the present application. Before step S103, the following may be included:

[0093] S10. Determine labeling information of multiple indicators corresponding to each sample data according to business experience to determine positive indicators and negative indicators, wherein the positive indicators have a positive impact on the user's loyalty score, and the negative indicators have a negative impact on the user's loyalty score.

[0094] For example, if the user is a bank depositor, and multiple indicators include deposits, transfers, loans, financial management, and consumption, then based on business experience, the information marked as deposits, transfers (transfers in), financial management (purchases), and loans (no overdue) is positive; the information marked as transfers (transfers out), financial management (sells), and loans (overdue) is negative. If the user is a mobile / telecom user, and multiple indicators include traffic and phone charges, then based on business experience, the information marked as traffic and phone charges (sufficient) is positive, and the information marked as phone charges (arrears) is negative.

[0095] S11. Bind the sample data corresponding to each user and the annotation information of multiple indicators to obtain the characteristic data of the corresponding user.

[0096] Specifically, the sample data corresponding to each user and the annotation information of multiple indicators can be bound and associated to generate a wide table, which is convenient for obtaining corresponding feature data when the entropy method is used to calculate the scores of each user.

[0097] S13, preprocessing the characteristic data of each user, wherein the preprocessing includes: processing the positive index and the negative index using different normalization methods.

[0098] In this embodiment, since the dimensions, magnitudes, and positive and negative orientations of the various indicators are different, in order to eliminate the impact of different dimensions on the score, it is necessary to standardize the various indicators, thereby solving the homogeneity problem of the values ​​of various different qualitative indicators, and at the same time, the negatively marked indicators are non-negatively processed, that is, the higher the positive indicator value, the better, and the lower the negative indicator value, the better.

[0099] Specifically, the characteristic data of each user is preprocessed, including: abnormal values, missing values, dimensions, magnitudes, and the positive and negative directions of indicators, etc., and normalized. Among them, abnormal values ​​in the data can be removed or corrected to normal values, and missing values ​​can be filled; the values ​​of positive indicators can be normalized using the following formula (6), and the values ​​of negative indicators can be normalized using the following formula (7). Specifically,

[0100] X′1=(x1-Xmin) / (Xmax-Xmin) (6)

[0101] Among them, x1 is the index value of the jth index of the i-th user when it is a positive index, and X′1 is the normalized index value. Xmax is the maximum value of the jth index of all users, and Xmin is the minimum value of the jth index of all users.

[0102] X′2=(Xmax-x2) / (Xmax-Xmin) (7)

[0103] Among them, x2 is the index value when the jth index of the i-th user is a negative index, and X′2 is the normalized index value.

[0104] Then step S103 specifically includes: for each user's characteristic data, using the entropy method to determine the proportion of each indicator in the characteristic data in the characteristic data of all users.

[0105] A possible implementation manner is also provided in the embodiment of the present application. Before S105, the method further includes:

[0106] S14. Standardize the multiple sample data to obtain a data set.

[0107] Specifically, in this embodiment, z_score standardization (zero-mean normalization, also known as standard deviation standardization) can be used to process multiple sample data. The mean of the processed data is 0 and the standard deviation is 1. The specific conversion formula is: in, is the mean of the original data, and υ is the standard deviation of the original data.

[0108] S15. Perform translation processing on the negative data in the data set to obtain non-negative data, wherein the translation amount during the translation processing is the sum of a preset value and the absolute value of the minimum data in the data set.

[0109] Specifically, in this embodiment, the preset value is used to ensure that the minimum data in the data set is not a non-negative value after translation processing, and can be 0.01, 0.001, etc. The specific conversion formula is: x=x_z+a, where xz is any negative data in the data set after standardization processing, x is the non-negative data after translation of xz, a is the translation amount, and a=|min(x_z)|+0.01, min(x_z) is the minimum data in the data set after standardization processing.

[0110] Specifically, in this embodiment, S105 may include: calculating the correlation coefficient matrix of each indicator in the standardized sample data, calculating the factor loading matrix, selecting common factors for scoring, performing factor rotation, and calculating the scores of individual common factors to obtain the scores of each scoring factor. S105 may include: using the variance contribution rate of each common factor as a weight, weighting the common factor score, and calculating the comprehensive factor score of each sample, that is, the second loyalty score of each user.

[0111] It should be noted that, in this embodiment, non-negative processing is performed on the negative data in the standardized data set, which can increase the sample size and make the scoring result obtained by the factor analysis method more accurate.

[0112] It should be understood that the steps not described in detail in the process of calculating the second loyalty score by using the factor analysis method can be implemented by using existing related technologies, and for the sake of brevity, they will not be repeated here.

[0113] A possible implementation manner is also provided in the embodiment of the present application. Before step S107, the method further includes:

[0114] S16. Perform a consistency check on the first loyalty scores and the second loyalty scores of all users corresponding to the plurality of sample data. If the consistency check passes, execute step S107.

[0115] Specifically, in this embodiment, step S16 may specifically include:

[0116] S1. Calculate the intraclass correlation coefficient ICC based on the first loyalty scores and the second loyalty scores of all users.

[0117] In this embodiment, the ICC may be calculated using existing related technologies, for example, a variance analysis method may be used, and specifically, the ICC may be calculated using an analysis of variance (ANNOVA) in SPSS software.

[0118] S2. If the intraclass correlation coefficient ICC of the first loyalty scores and the second loyalty scores of all users is within a preset range, it is determined that the first loyalty scores and the second loyalty scores of all users pass the consistency check.

[0119] Specifically, in this embodiment, if the first loyalty scores and the second loyalty scores of all users corresponding to the multiple sample data pass the consistency check, the average of the first loyalty score and the second loyalty score of each user can be used as the final score for evaluating user loyalty, that is, the target score of each user.

[0120] The preset range is between 0.5 and 1. ICC is equal to the individual variability divided by the total variability, so its value is between 0 and 1, where 0 means unreliable and 1 means completely reliable. It is generally believed that less than 0.4 indicates poor reliability, and greater than 0.75 indicates good reliability.

[0121] That is to say, in this embodiment, assuming that there are 1000 sample data, that is, data of 1000 users, the entropy method is used to obtain the corresponding 1000 first loyalty scores, and the factor analysis method is used to obtain the corresponding 1000 second loyalty scores, and then the ICC is calculated based on the 1000 first loyalty scores and the 1000 second loyalty scores. If the ICC is greater than 0.5, it is determined that the consistency check has passed, and then, for each user, the first loyalty score and the second loyalty score are aggregated (for example: the average is calculated) to obtain the final score of the corresponding user.

[0122] The intraclass correlation coefficient ICC is usually used to evaluate the consistency or reliability of different measurement methods or assessors for unified quantitative measurement results. After using multiple algorithms for scoring, it is necessary to consider whether the scoring results of different scoring methods for the same sample are additively correlated. Therefore, the present invention can determine whether the two algorithms can be aggregated into higher-order data by testing the ICC values ​​of the scoring results of each sample by the entropy method and factor analysis, and then analyze them, such as finding the mean.

[0123] Therefore, in the embodiment of the present application, the user loyalty score is scored by using factor analysis and entropy method respectively, and when the two scoring results are consistent, the two scoring results are aggregated to obtain the user loyalty score, thereby making the obtained user loyalty score more accurate.

[0124] The present application embodiment provides a user loyalty scoring device, such as Figure 2 As shown, the user loyalty scoring device may include: an acquisition module 201, a pre-processing module 202 and a determination module 203, wherein:

[0125] An acquisition module 201 is used to acquire a plurality of sample data, each of which corresponds to a user and includes a plurality of parameters;

[0126] The preprocessing module 202 is used to perform preset processing on each parameter in the plurality of sample data to obtain an index for evaluating user loyalty; wherein the preset processing includes: parameter screening and parameter derivation;

[0127] Determination module 203 is used to determine, for each sample data, the proportion of each indicator in the sample data in the multiple sample data by using the entropy method; it is also used to determine, for each user, the first loyalty score of the user based on the proportion of each indicator in the sample data corresponding to the user and the weight of each indicator; it is also used to determine, by using the factor analysis method, the correlation between each indicator in the multiple sample data, determine each scoring factor and the score of each scoring factor corresponding to each sample data; it is also used to determine, for each user, the second loyalty score of the user based on the score of each scoring factor corresponding to the user; for each user, determine the target score of the user based on the first loyalty score and the second loyalty score of the user.

[0128] The user loyalty scoring device provided in the embodiment of the present application specifically implements the above method embodiment process. For details, please refer to the content of the above user loyalty scoring method embodiment, which will not be repeated here. The user loyalty scoring device provided in the embodiment of the present application uses factor analysis and entropy method to score user loyalty respectively, and obtains the user's loyalty score based on the two scores, thereby solving the problem of low accuracy of user loyalty scoring.

[0129] In a possible implementation, when the determination module 203 determines the second loyalty score of each user based on the scores of the scoring factors corresponding to the user, the determination module 203 is specifically configured to:

[0130] For each user, the variance contribution rate of each scoring factor corresponding to the user is used as the weight of each scoring factor;

[0131] The scores of the scoring factors are weighted and calculated to obtain the second loyalty score of the user.

[0132] In another possible implementation, the determining module 203 is further configured to:

[0133] Determine the labeling information of multiple indicators corresponding to each sample data according to business experience to determine positive indicators and negative indicators, wherein the positive indicators have a positive impact on the user's loyalty score, and the negative indicators have a negative impact on the user's loyalty score;

[0134] The sample data corresponding to each user is bound to the annotation information of multiple indicators to obtain the characteristic data of the corresponding user.

[0135] The preprocessing module 202 is further used to preprocess the characteristic data of each user, wherein the preprocessing includes processing the positive index and the negative index using different normalization methods.

[0136] When the determination module 203 uses the entropy method to determine the proportion of each indicator in the sample data in the plurality of sample data for each sample data, it is specifically used to:

[0137] For each user's characteristic data, an entropy method is used to determine the proportion of each indicator in the characteristic data in the characteristic data of all users.

[0138] In another possible implementation, the preprocessing module 202 is further used to: perform standardization on the multiple sample data to obtain a data set; perform translation on the negative data in the data set to obtain non-negative data, wherein the translation amount during the translation is the sum of a preset value and the absolute value of the minimum data in the data set.

[0139] In yet another possible implementation, the determining module 203 is further configured to:

[0140] Performing consistency check on the first loyalty scores and the second loyalty scores of all users corresponding to the plurality of sample data;

[0141] When the consistency check passes, for each user, a target score of the user is determined based on the first loyalty score and the second loyalty score of the user.

[0142] In yet another possible implementation, when performing consistency check on the first loyalty scores and the second loyalty scores of all users corresponding to the plurality of sample data, the determination module 203 is specifically configured to:

[0143] Calculate the intraclass correlation coefficient ICC based on the first loyalty score and the second loyalty score of all users;

[0144] If the intraclass correlation coefficient ICC is within a preset range, it is determined that the first loyalty scores and the second loyalty scores of all users pass the consistency check.

[0145] An electronic device is provided in an embodiment of the present application, including a memory, a processor, and a computer program stored in the memory. The processor executes the above-mentioned computer program to implement the steps of a user loyalty scoring method. Compared with the related art, the following can be achieved: In the embodiment of the present application, user loyalty scoring is performed by using factor analysis method and entropy method respectively, and the user's loyalty score is obtained based on the two scores, thereby solving the problem of low accuracy of user loyalty scoring.

[0146] In an alternative embodiment, an electronic device is provided, such as Figure 3 As shown, Figure 3 The electronic device 300 shown includes: a processor 301 and a memory 303. The processor 301 and the memory 303 are connected, such as through a bus 302. Optionally, the electronic device 300 may also include a transceiver 304, which may be used for data interaction between the electronic device and other electronic devices, such as data transmission and / or data reception. It should be noted that in actual applications, the transceiver 304 is not limited to one, and the structure of the electronic device 300 does not constitute a limitation on the embodiments of the present application.

[0147] Processor 301 may be a CPU (Central Processing Unit), a general purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. Processor 301 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0148] The bus 302 may include a path to transmit information between the above components. The bus 302 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 302 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0149] The memory 303 can be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), magnetic disk storage medium, other magnetic storage devices, or any other medium that can be used to carry or store computer programs and can be read by a computer, without limitation herein.

[0150] The memory 303 is used to store the computer program for executing the embodiment of the present application, and the execution is controlled by the processor 301. The processor 301 is used to execute the computer program stored in the memory 303 to implement the steps shown in the above method embodiment.

[0151] An embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps and corresponding contents of the aforementioned method embodiment can be implemented.

[0152] The terms "first", "second", "third", "fourth", "1", "2", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than that shown or described in the drawings.

[0153] It should be understood that, although each operation step is indicated by arrows in the flowchart of the embodiment of the present application, the implementation order of these steps is not limited to the order indicated by the arrows. Unless clearly stated herein, in some implementation scenarios of the embodiment of the present application, the implementation steps in each flowchart can be performed in other orders according to demand. In addition, some or all of the steps in each flowchart may include multiple sub-steps or multiple stages based on actual implementation scenarios. Some or all of these sub-steps or stages may be executed at the same time, and each sub-step or stage in these sub-steps or stages may also be executed at different times respectively. In different scenarios of execution time, the execution order of these sub-steps or stages may be flexibly configured according to demand, and the embodiment of the present application does not limit this.

[0154] The above is only an optional implementation method for some implementation scenarios of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the technical concept of the solution of the present application, other similar implementation methods based on the technical ideas of the present application are also within the protection scope of the embodiments of the present application.

Claims

1. A user loyalty scoring method, characterized in that: The method comprises: Acquire multiple sample data, each sample data corresponds to a user, and each sample data includes multiple parameters; Performing preset processing on each parameter in the plurality of sample data to obtain an index for evaluating user loyalty; wherein the preset processing includes: parameter screening and parameter derivation; For each sample data, an entropy method is used to determine the proportion of each indicator in the sample data in the multiple sample data; For each user, based on the proportion of each indicator of the sample data corresponding to the user and the weight of each indicator, determine a first loyalty score of the user; Using factor analysis to determine the correlation between the indicators in the plurality of sample data, and to determine the scoring factors and the scores of the scoring factors corresponding to the sample data; For each user, determining a second loyalty score of the user based on the scores of the scoring factors corresponding to the user; Performing consistency check on the first loyalty scores and the second loyalty scores of all users corresponding to the plurality of sample data; When the consistency check passes, for each user, the first loyalty score and the second loyalty score of the user are aggregated to determine a target score of the user.

2. The user loyalty scoring method according to claim 1, characterized in that: Before the step of using factor analysis to determine the correlation between the indicators in the plurality of sample data and determining the scoring factors and the scores of the scoring factors corresponding to the sample data, the method further includes: Performing standardization processing on the plurality of sample data to obtain a data set; The negative data in the data set are subjected to translation processing to obtain non-negative data, wherein the translation amount during the translation processing is the sum of a preset value and the absolute value of the minimum data in the data set.

3. The user loyalty scoring method according to claim 1, characterized in that: The step of performing consistency verification on the first loyalty scores and the second loyalty scores of all users corresponding to the plurality of sample data includes: Based on the first loyalty score and the second loyalty score of all users, the intraclass correlation coefficient ICC is calculated; If the intraclass correlation coefficient ICC is within a preset range, it is determined that the first loyalty scores and the second loyalty scores of all users pass the consistency check.

4. The user loyalty scoring method according to claim 1, characterized in that: The step of determining, for each user, the second loyalty score of the user based on the scores of the scoring factors corresponding to the user, comprises: For each user, the variance contribution rate of each scoring factor corresponding to the user is used as the weight of each scoring factor; The scores of the scoring factors are weighted and calculated to obtain the second loyalty score of the user.

5. The user loyalty scoring method according to claim 1, characterized in that: Before the step of using the entropy method to determine the proportion of each indicator in the sample data in the plurality of sample data for each sample data, the method further includes: Determine the labeling information of multiple indicators corresponding to each sample data according to business experience to determine positive indicators and negative indicators, wherein the positive indicators have a positive impact on the user's loyalty score, and the negative indicators have a negative impact on the user's loyalty score; Bind the sample data corresponding to each user and the annotation information of multiple indicators to obtain the characteristic data of the corresponding user; Preprocessing is performed on the characteristic data of each user, and the preprocessing includes: using different normalization methods to process positive indicators and negative indicators; The step of using the entropy method to determine the proportion of each indicator in the sample data in the plurality of sample data for each sample data includes: For each user's characteristic data, an entropy method is used to determine the proportion of each indicator in the characteristic data in the characteristic data of all users.

6. A user loyalty scoring device, characterized in that: The device comprises: An acquisition module is used to acquire multiple sample data, each sample data corresponds to a user, and each sample data includes multiple parameters; A preprocessing module, used to perform preset processing on each parameter in the plurality of sample data to obtain an index for evaluating user loyalty; wherein the preset processing includes: parameter screening and parameter derivation; A determination module is used to determine, for each sample data, the proportion of each indicator in the sample data in the multiple sample data by using an entropy method; it is also used to determine, for each user, the first loyalty score of the user based on the proportion of each indicator in the sample data corresponding to the user and the weight of each indicator; it is also used to determine, by using a factor analysis method, the correlation between each indicator in the multiple sample data, determine each scoring factor and the score of each scoring factor corresponding to each sample data; it is also used to determine, for each user, the second loyalty score of the user based on the score of each scoring factor corresponding to the user; a consistency check is performed on the first loyalty score and the second loyalty score of all users corresponding to the multiple sample data; if the consistency check passes, for each user, the first loyalty score and the second loyalty score of the user are aggregated to determine the target score of the user.

7. The user loyalty scoring device according to claim 6, characterized in that: The preprocessing module is further used for: Performing standardization processing on the plurality of sample data to obtain a data set; The negative data in the data set are subjected to translation processing to obtain non-negative data, wherein the translation amount during the translation processing is the sum of a preset value and the absolute value of the minimum data in the data set.

8. An electronic device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the user loyalty scoring method according to any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the user loyalty scoring method described in any one of claims 1 to 5 are implemented.

Images