Downshift user identification method and device

Abstract

The invention provides a downshift user identification method and device, and relates to the technical field of communication. The method comprises the following steps: acquiring respective package change record data sets of a plurality of first users, and performing feature extraction on each package change record data set to obtain respective first data sets of the plurality of first users; screening out a plurality of second data sets from the plurality of first data sets through a preset downshift identification model; calculating the downshift probability of each second data set through a preset artificial intelligence algorithm; according to the downshift probability of each second data set and the obtained first number of downshift users, obtaining a probability threshold value, so as to judge whether the package grade of the user is reduced or not according to the probability threshold value; wherein the first number of downshift users refers to the actual number of users with reduced package grade in the plurality of first users. According to the method provided by the invention, the problem of how to accurately identify the user with the reduced package grade is solved.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a downshift user identification method and device. Background Art

[0002] Telecom operators often package different types of services into "packages" for sale. The packages usually include multiple resources such as voice, data, and text messages under one or more user numbers, and give users more favorable package prices. Therefore, users can downgrade the package level or open a new package to achieve the discount.

[0003] Existing methods for predicting the downgrade of user package levels include: screening through simple rule-based classification, screening variables through correlation analysis, and predicting through model building. Although such methods can predict the user's package reduction, it is difficult to improve the accuracy of the prediction through simple rule-based judgment, and the use of analytical methods requires a large amount of data and cannot accurately predict.

[0004] Therefore, how to accurately identify users who need to downgrade their package levels is an urgent problem to be solved in the present invention. Summary of the invention

[0005] The present application provides a method and device for identifying downgraded users, so as to solve the problem of how to accurately identify users who downgrade their package levels.

[0006] In a first aspect, the present application provides a downshift user identification method, the method comprising:

[0007] Acquire package change record data sets of respective multiple first users, and perform feature extraction on each package change record data set respectively to obtain first data sets of respective multiple first users; wherein the first data set refers to a data set related to the package level;

[0008] Filtering a plurality of second data sets from a plurality of first data sets by using a preset downshift recognition model;

[0009] The downgrade probability of each second data set is calculated by a preset artificial intelligence algorithm; wherein the downgrade probability refers to the probability of the user's package level being downgraded;

[0010] A probability threshold is obtained based on the downgrade probability of each second data set and the first number of downgraded users obtained, so as to determine whether the package level of the user is downgraded based on the probability threshold; wherein the first number of downgraded users refers to the actual number of users whose package levels are downgraded among multiple first users.

[0011] In a possible design, the plurality of downshift probabilities are all within a first value range;

[0012] According to the downshift probability of each second data set and the acquired first number of downshifted users, a probability threshold is obtained, including:

[0013] According to a plurality of preset division thresholds, the first numerical range is divided into a plurality of second numerical ranges; wherein the minimum value of each second numerical range is equal to a division threshold, and the maximum value of each second numerical range is equal to the maximum value of the first numerical range;

[0014] According to the multiple downshift probabilities, the accumulated number of users for each division threshold is obtained by summarizing; wherein the accumulated number of users refers to the number of downshift probabilities within the corresponding second numerical range;

[0015] A probability threshold is determined from a plurality of partitioning thresholds according to the accumulated number of users of each partitioning threshold and the first number of downshifted users.

[0016] In a possible design, according to the accumulated number of users of each partition threshold and the first number of downgraded users, a probability threshold is determined from multiple partition thresholds, including:

[0017] Calculate the difference between the accumulated number of each user and the first number of downshifted users;

[0018] The division threshold with the smallest absolute value of the quantity difference is determined as the probability threshold.

[0019] In a possible design, each second data set includes a plurality of characteristic data, and the characteristic data refers to package-level related data;

[0020] The downshift probability of each second data set is calculated by a preset artificial intelligence algorithm, including:

[0021] Performing weighted calculation on the plurality of characteristic data of each second data set to obtain a weighted calculation result of each second data set;

[0022] By using a preset mapping function, a plurality of weighted calculation results are mapped to a first numerical range to obtain a mapping result of each weighted calculation result;

[0023] The mapping result of each weighted calculation result is determined as the downshift probability of the corresponding second data set.

[0024] In a possible design, each first data set includes: primary and secondary card identification information, card activation record information, package level information, and package value information;

[0025] The second user is any one of the plurality of first users. For the second user, a plurality of second data sets are screened out from the plurality of first data sets by using a preset downshift recognition model, including:

[0026] When the downgrade identification model determines that the second user's previous number is switched from a main card to a secondary card of a newly opened number, the current package level of the newly opened number is lower than the original package level of the previous number, and the current package value of the second user is lower than the original package level of the second user, the first data set of the second user is determined as the second data set.

[0027] In a possible design, before extracting features from each package change record data set to obtain respective first data sets of multiple first users, the method further includes:

[0028] Perform data cleaning on each package change record dataset;

[0029] Each package change record data set after data cleaning is standardized.

[0030] In a second aspect, the present application provides a downshift user identification method, the method comprising:

[0031] Each third user whose downgrade probability is greater than a preset probability threshold is determined as a user whose package level is downgraded; wherein the probability threshold is a threshold obtained by the downgrade user identification method of the first aspect of the invention.

[0032] In a possible design, after determining each third user whose downgrade probability is greater than a preset probability threshold as a user whose package level is downgraded, the method further includes:

[0033] Acquire a second number of downgraded users; wherein the second number of downgraded users refers to the actual number of users whose package levels are downgraded among the plurality of third users;

[0034] The probability threshold is updated according to the downshift probabilities of the plurality of third users and the second number of downshifted users.

[0035] In a third aspect, the present application provides a downshift user identification device, comprising:

[0036] A data acquisition module is used to acquire package change record data sets of multiple first users, and perform feature extraction on each package change record data set to obtain first data sets of multiple first users; wherein the first data set refers to a data set related to the package level;

[0037] A model screening module, used to screen out a plurality of second data sets from a plurality of first data sets by using a preset downshift recognition model;

[0038] A probability calculation module, used to calculate the downgrade probability of each second data set through a preset artificial intelligence algorithm; wherein the downgrade probability refers to the probability that the package level of the user is downgraded;

[0039] The judgment module is used to obtain a probability threshold according to the downgrade probability of each second data set and the first number of downgraded users obtained, so as to judge whether the package level of the user is reduced according to the probability threshold; wherein the first number of downgraded users refers to the actual number of users whose package levels are reduced among multiple first users.

[0040] In a fourth aspect, the present application provides a downshift user identification device, comprising:

[0041] An acquisition probability module, used to acquire the downshift probability of each of a plurality of third users;

[0042] The user screening module is used to determine each third user whose downgrade probability is greater than a preset probability threshold as a user whose package level is downgraded; wherein the probability threshold is a threshold obtained by the downgrade user identification device of the third aspect of the invention.

[0043] In a fifth aspect, the present application provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor;

[0044] Memory stores computer-executable instructions;

[0045] The processor executes the computer-executable instructions stored in the memory to implement a downshift user identification method according to the first invention content and the second invention content.

[0046] In a sixth aspect, the present application provides a computer-readable storage medium, in which computer execution instructions are stored. When the computer execution instructions are executed by a processor, a downshift user identification method is used to implement the first aspect of the invention and the second aspect of the invention.

[0047] In a seventh aspect, the present application provides a computer program product, including: a computer program, which, when executed by a processor, is used to implement a downshift user identification method according to the first aspect of the invention and the second aspect of the invention.

[0048] The present application provides a method and device for identifying downgraded users, which obtains package change record data sets of multiple first users, and extracts features from each package change record data set to obtain first data sets of multiple first users; screens multiple second data sets from multiple first data sets through a preset downgrade identification model; calculates the downgrade probability of each second data set through a preset artificial intelligence algorithm; obtains a probability threshold based on the downgrade probability of each second data set and the first number of downgraded users obtained, so as to judge whether the user's package level is reduced according to the probability threshold; wherein the first number of downgraded users refers to the actual number of users whose package level is reduced among multiple first users. The following technical effects are achieved: by extracting features from data and screening user data, a large amount of data is removed, so as to timely screen out downgraded users; through the preset downgrade identification model, the model can receive new user package data and output the evaluation result of downgrade, providing strong support for risk warning; the downgrade probability is calculated according to the artificial intelligence algorithm, and the key features of the model and important parameters such as thresholds are optimized, thereby improving the accuracy and adaptability of prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 A schematic diagram of the system architecture of the downshift user identification method provided in an embodiment of the present application;

[0051] Figure 2 Schematic diagram of the process of downshift user identification method provided in the embodiment of the present application Figure 1 ;

[0052] Figure 3 Schematic diagram of the process of downshift user identification method provided in the embodiment of the present application Figure 2 ;

[0053] Figure 4 Schematic diagram of the process of downshift user identification method provided in the embodiment of the present application Figure 3 ;

[0054] Figure 5 Schematic diagram of the process of downshift user identification method provided in the embodiment of the present application Figure 4 ;

[0055] Figure 6 A schematic diagram of the structure of the downshift user identification device provided in the embodiment of the present application Figure 1 ;

[0056] Figure 7 A schematic diagram of the structure of the downshift user identification device provided in the embodiment of the present application Figure 2 ;

[0057] Figure 8 A schematic diagram of the structure of the electronic device hardware provided in an embodiment of the present application.

[0058] Reference numerals:

[0059] 100-downshift user identification server; 110-model evaluation module; 120-probability prediction module; 130-real-time update module;

[0060] 600 - downshift user identification device; 610 - data acquisition module; 620 - model screening module; 630 - probability calculation module; 640 - judgment module;

[0061] 710-acquisition probability module; 720-user screening module;

[0062] 800 - electronic device; 810 - processor; 820 - memory; 830 - communication component; 840 - bus. DETAILED DESCRIPTION

[0063] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0064] In the embodiments of the present application, words such as "first" and "second" are used to distinguish the same or similar items with basically the same functions and effects. Those skilled in the art will understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit the difference. It should be noted that in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" or "for example" in the present application should not be interpreted as being more preferred or more advantageous than other embodiments or design schemes. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way. In the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more.

[0065] It should be noted that the "at..." in the embodiment of the present application can be the instant when a certain situation occurs, or can be a period of time after a certain situation occurs, and the embodiment of the present application does not specifically limit this. In addition, the downshift user identification method provided in the embodiment of the present application is only an example, and the downshift user identification method can also include more or less content.

[0066] In order to clearly describe the technical solutions of the embodiments of the present application, some terms and technologies involved in the embodiments of the present application are briefly introduced below:

[0067] Downgrading: usually refers to a user choosing a package plan with lower costs and fewer resources when using mobile communications, Internet services or other subscription services. In this application, it is referred to as downgrading and is only applied to mobile communications services.

[0068] Feature extraction: Feature extraction is an important step in data processing and machine learning, which aims to extract useful and representative information from raw data for subsequent analysis or modeling. The goal of feature extraction is to convert high-dimensional and complex data into a low-dimensional and concise form while retaining the key information of the data.

[0069] Primary Card: The Primary Card is the main SIM (Subscriber Identity Module) card provided by the mobile operator to the account holder.

[0070] Secondary card: A secondary card is an additional SIM card applied for by the primary card holder for other family members or colleagues. The secondary card shares the same package resources with the primary card.

[0071] Package level: refers to the different levels of service packages provided by mobile communication operators. Each level contains different resources and services, and usually corresponds to different prices.

[0072] Artificial Intelligence Algorithms: Artificial Intelligence (AI) algorithms are a class of technologies and methods used to simulate human intelligence. They enable computer systems to perform complex tasks such as learning, decision-making, and perception. AI algorithms have a wide range of applications in many fields, including natural language processing and image processing.

[0073] A package usually contains one or more numbers, and multiple numbers share one package. Currently, users seek more favorable prices for mobile packages, and in the market, in order to achieve marketing development goals, users are also guided to apply for a new package with a lower level than the original main card, and the original main card is downgraded to the secondary card of the new card, and sold together to users.

[0074] Therefore, in order to identify this behavior, the implementation method of predicting the downgrade of user package level includes:

[0075] First, the correlation analysis method is used to screen the variables required for downsizing to obtain the data set;

[0076] Secondly, the data set is input into the downsizing risk prediction model;

[0077] Finally, the downsizing risk is predicted through the downsizing risk model.

[0078] Or it can be used to identify risky behaviors through simple rules. However, there may be some technical problems in using such methods to predict the downsizing behavior:

[0079] First, through simple rule judgment, downgrade behavior identification can be achieved. Risk data that meets the rules can be extracted, but the amount of extracted data is too large and the accuracy is low, so it is impossible to accurately locate the real problem, and the operation is relatively complicated;

[0080] Secondly, the preset downsizing risk prediction model was used to predict downsizing behavior, but the amount of data used was also large, and the accuracy of the prediction was not improved; and the prediction results could not be updated based on real-time data, so the accuracy and applicability were not high.

[0081] Therefore, how to accurately identify users who need to downgrade their package levels is an urgent problem to be solved in the present invention.

[0082] Based on this, an embodiment of the present application provides a downshift user identification method and device, which can be used in the field of communication technology, aiming to solve the above technical problems of the prior art.

[0083] Figure 1 This is a schematic diagram of the system architecture of the downshift user identification method provided in the embodiment of the present application. It should be noted that: Figure 1 What is shown is merely an example of a system architecture to which the embodiments of the present application can be applied, in order to help those skilled in the art understand the technical content of the present application, but it does not mean that the embodiments of the present application cannot be used in other devices, systems, environments or scenarios.

[0084] like Figure 1 As shown, the system architecture of the downshift user identification method includes: the downshift user identification server 100 includes a model evaluation module 110, a probability prediction module 120 and a real-time update module 130.

[0085] The model evaluation module 110 is used to input the data set after feature extraction and output it through the model.

[0086] The probability prediction module 120 is used to process the data set output by the model evaluation module 110, and calculate the probability threshold corresponding to the user data through an artificial intelligence algorithm, so as to filter the user data according to the probability threshold to obtain a prediction result.

[0087] The real-time update module 130 is used to update and calculate the probability threshold of each package change data based on the acquired real-time data, so as to take prevention and control measures in time.

[0088] Figure 2 Schematic diagram of the process of downshift user identification method provided in the embodiment of the present application Figure 1 .like Figure 2 As shown, the method includes:

[0089] S201. Acquire package change record data sets of respective multiple first users, and perform feature extraction on each package change record data set to obtain first data sets of respective multiple first users.

[0090] The first data set refers to the data set related to the package level.

[0091] Specifically, a plurality of first users' respective package change record data sets are obtained, and data cleaning, integration and standardization are performed on each first user's respective package change record data set. In addition, according to business rules and data characteristics, combined with business experience, feature extraction is performed from each package change record data set to obtain the first data sets of the plurality of first users.

[0092] S202, screening out a plurality of second data sets from a plurality of first data sets by using a preset downshift recognition model;

[0093] Specifically, in this application, the downgrade identification model includes four rules, namely: the previous number changes from the main card to the secondary card, the newly opened number is the main card of the previous number, the package level of the new number is lowered relative to the original package level of the previous number, and the package value of the new number is lower than the original package value of the previous number.

[0094] Through the preset downshift recognition model, multiple second data sets are obtained by screening from multiple first data sets according to four preset rules, wherein the screened second data sets must satisfy the four preset rules at the same time.

[0095] S203: Calculate the downshift probability of each second data set by using a preset artificial intelligence algorithm.

[0096] Among them, the downgrade probability refers to the probability that the user's package level is reduced.

[0097] Specifically, in this application, the machine learning algorithm in the artificial intelligence algorithm realizes the calculation of probability through the method of data learning patterns and rules.

[0098] Through the preset artificial intelligence algorithm, each data corresponding to each second data set is subjected to a weighted sum algorithm to obtain a calculation result corresponding to each second data set. Then, each calculation result is mapped to a numerical range of 0-1 through a function and a corresponding artificial intelligence algorithm to obtain a mapping result of each second data set, that is, a corresponding downshift probability.

[0099] S204: Obtain a probability threshold according to the downgrade probability of each second data set and the acquired first number of downgraded users, so as to determine whether the package level of the user is downgraded according to the probability threshold.

[0100] The first number of downgraded users refers to the actual number of users whose package levels are downgraded among multiple first users.

[0101] Specifically, the downshift probability corresponding to each second data set is divided in a regular manner so that each probability is increased in a recursive manner, thereby obtaining a plurality of division thresholds.

[0102] Obtain the first number of downgraded users, and according to each division threshold, compare the obtained first number of downgraded users with the number of downgraded users at each division threshold, select the probability threshold closest to the number, and determine the probability threshold so as to judge the package level of the user according to the probability threshold.

[0103] The present application provides a method and device for identifying downgraded users, which obtains package change record data sets of multiple first users, and extracts features from each package change record data set to obtain first data sets of multiple first users; screens multiple second data sets from multiple first data sets through a preset downgrade identification model; calculates the downgrade probability of each second data set through a preset artificial intelligence algorithm; obtains a probability threshold based on the downgrade probability of each second data set and the first number of downgraded users obtained, so as to judge whether the user's package level is reduced according to the probability threshold; wherein the first number of downgraded users refers to the actual number of users whose package level is reduced among multiple first users. The following technical effects are achieved: by extracting features from data and screening user data, a large amount of data is removed, so as to timely screen out downgraded users; through the preset downgrade identification model, the model can receive new user package data and output the evaluation result of downgrade, providing strong support for risk warning; the downgrade probability is calculated according to the artificial intelligence algorithm, and the key features of the model and important parameters such as thresholds are optimized, thereby improving the accuracy and adaptability of prediction.

[0104] Figure 3 Schematic diagram of the process of downshift user identification method provided in the embodiment of the present application Figure 2 In this embodiment Figure 2 Based on the embodiment, the downshift user identification method is described in detail, such as Figure 3 As shown, this embodiment provides a downshift user identification method, including:

[0105] S301. Obtain package change record data sets of multiple first users.

[0106] The principles of S301 and S201 are similar and will not be described in detail in this embodiment.

[0107] S302: Perform data cleaning on each package change record data set.

[0108] Specifically, the package change records corresponding to each first user are cleaned to remove incomplete, erroneous or inconsistent data, and then integrated to keep the data format consistent.

[0109] S303, standardize each package change record data set after data cleaning.

[0110] Specifically, each cleaned package change record is standardized and the data is converted into a unified format or range.

[0111] S304: Perform feature extraction on each package change record data set to obtain respective first data sets of multiple first users.

[0112] The first data set refers to the data set related to the package level.

[0113] Specifically, according to business rules and data characteristics, combined with business experience, feature extraction is performed from each package change record data set to screen out the first data sets of multiple first users, which are the inputs of the subsequent downshift recognition model.

[0114] S305. When the downgrade recognition model determines that the previous number of the second user is switched from a main card to a secondary card of a newly opened number, the current package level of the newly opened number is lower than the original package level of the previous number, and the current package value of the second user is lower than the original package level of the second user, the first data set of the second user is determined as the second data set.

[0115] Each first data set includes: primary and secondary card identification information, card activation record information, package level information and package value information; the second user is any one of the multiple first users.

[0116] Specifically, multiple second data sets are screened out from multiple first data sets through a preset downgrade recognition model. When the downgrade recognition model determines that the primary card of the second user's past number is downgraded to a secondary card, and it is a secondary card of a newly opened number; the current package level of the card of the newly opened number is lower than the original package level of the card of the original past number, and the current package value of the existing second user is lower than the package value of the original past number, the first data set of the second user is determined as the second data set.

[0117] S306: Perform weighted calculation on the plurality of characteristic data of each second data set to obtain a weighted calculation result of each second data set.

[0118] Each second data set includes a plurality of characteristic data, where the characteristic data refers to data related to the package level.

[0119] Specifically, the downgrade probability of each second data set is calculated by a preset artificial intelligence algorithm, wherein the downgrade probability refers to the probability of the user's package level being reduced. The weighted calculation result of each second data set is obtained by calculating the multiple feature data of each second data set by a weighted algorithm.

[0120] S307: Map multiple weighted calculation results to a first numerical range through a preset mapping function to obtain a mapping result of each weighted calculation result.

[0121] Specifically, the preset function may include a sigmoid function in a logistic regression algorithm, and the logistic regression algorithm may convert an input variable into a probability value between 0 and 1.

[0122] The weighted calculation result of each second data set is mapped to the first numerical unit through a preset function to obtain a mapping result of each weighted calculation result.

[0123] That is, by using the logistic regression algorithm, the weighted calculation result of each second data set is mapped to a numerical range of 0 to 1, and the probability value corresponding to each second data set is obtained.

[0124] S308: Determine the mapping result of each weighted calculation result as the downshift probability of the corresponding second data set.

[0125] Specifically, the mapping result of each weighted calculation result, that is, each probability value corresponding to the second data set, is determined as the downshift probability of the second data set corresponding to each second user.

[0126] S309: Divide the first numerical range into multiple second numerical ranges according to multiple preset division thresholds.

[0127] The minimum value of each second numerical range is equal to a division threshold, and the maximum value of each second numerical range is equal to the maximum value of the first numerical range. The multiple downshift probabilities are all within the first numerical range.

[0128] Specifically, a probability threshold is obtained based on the downgrade probability of each second data set and the first number of downgraded users obtained, so as to determine whether the package level of the user is downgraded based on the probability threshold; wherein the first number of downgraded users refers to the actual number of users whose package levels are downgraded among the multiple first users. The first numerical range is divided into multiple second numerical ranges based on a preset division threshold.

[0129] S310 . Summarize and obtain the cumulative number of users for each division threshold according to multiple downshift probabilities.

[0130] The user accumulated number refers to the number of downshift probabilities within the corresponding second numerical range.

[0131] Specifically, according to the multiple downshift probabilities, the number of second users corresponding to each second user in the corresponding division threshold is obtained.

[0132] S311, calculating the difference between the accumulated number of each user and the first number of downgraded users.

[0133] Specifically, according to the accumulated number of users of each division threshold and the first number of downgraded users, a probability threshold is determined from multiple division thresholds, and the difference between the accumulated number of each user and the actually acquired first number of downgraded users is calculated.

[0134] S312. Determine the partition threshold with the smallest absolute value of the quantity difference as the probability threshold.

[0135] Specifically, the partition threshold with the smallest absolute value of the difference between the first number of users and the number of each second user within the partition threshold is determined as the probability threshold.

[0136] For the convenience of explaining this embodiment, a specific embodiment is introduced here:

[0137] Assuming that there are 10 first users, there are 10 package change record data sets for the first users. The package change record data set includes the following data: current package details (package type, package level, and primary and secondary card identification, etc.), consumption (voice, traffic, SMS usage and fees, etc.), package change records (change time, change reason, handling channel, handling personnel, package type and level after change, and changes in primary and secondary card identification, etc.), and new card opening records (account opening time, selected package, activation status, customer ID information, sales channel and sales personnel, etc.).

[0138] The data sets of 10 package change records were cleaned, integrated and standardized.

[0139] According to business rules and data characteristics, and combined with previous business experience, 10 standardized package change record data sets are subjected to feature extraction to obtain the first data sets corresponding to 10 first users. The first data set includes the following data: primary and secondary card identification, new card opening records, package level, customer certificate information, handling channels (sales channels) and consumption status.

[0140] According to the preset downgrade identification model, the first data set corresponding to the 10 first users is screened and processed, wherein the preset downgrade identification model needs to meet the following four conditions: the previous number changes from a main card to a secondary card, the newly opened number is the main card of the previous number, the package level of the newly opened number is lower than the original package level of the previous number, and the package value of the newly opened number is lower than the original package value of the previous number.

[0141] After screening by the downshift identification model, it is obtained that there are 8 second users, and the probabilities corresponding to the second data sets of the 8 second users are 0.5, 0.4, 0.65, 0.7, 0.75, 0.6, 0.8 and 0.85 respectively.

[0142] Assuming that the division threshold is 0.1, the downshift probabilities corresponding to the second users are respectively used as the second numerical ranges, and the second numerical ranges are respectively above 0.4, above 0.5, above 0.6, above 0.7, and above 0.8. And the first number of downshift users obtained is 6.

[0143] There are 8 users with a score above 0.4, 7 users with a score above 0.5, 6 users with a score above 0.6, 3 users with a score above 0.7, and 1 user with a score above 0.8.

[0144] The number of users is compared with the first number of downshifted users, and when it is determined that the second numerical range is above 0.6, it is closest to the actual data, and the probability 0.6 corresponding to the second data set at this time is determined as the probability threshold.

[0145] The present application provides a method and device for identifying downgraded users, which obtain package change record data sets of multiple first users, and perform feature extraction on each package change record data set to obtain first data sets of multiple first users; screen out multiple second data sets from the multiple first data sets through a preset downgrade identification model; calculate the downgrade probability of each second data set through a preset artificial intelligence algorithm; obtain a probability threshold based on the downgrade probability of each second data set and the obtained first number of downgraded users, so as to determine whether the user's package level is reduced based on the probability threshold; wherein the first number of downgraded users refers to the actual number of users whose package levels are reduced among the multiple first users. The following technical effects are achieved: by extracting features from data and screening user data, a large amount of data is removed, so that users who are downgrading can be screened out in a timely manner; through the preset downgrade identification model, the model can receive new user package data and output the evaluation results of the downgrade, providing strong support for risk warning; the downgrade probability is calculated according to the artificial intelligence algorithm, and important parameters such as the key features and thresholds of the model are optimized to improve the accuracy and adaptability of the prediction; the data output by the downshift identification model is further improved in accuracy and combined with actual data to make the probability threshold more accurate; the acquired user first data set is cleaned and standardized to make the data more concise, so that the amount of data adopted is smaller, and it is easy to accurately locate the problem.

[0146] Figure 4 Schematic diagram of the process of downshift user identification method provided in the embodiment of the present application Figure 3 In this embodiment Figure 2 Based on the embodiment, a downshift user identification method is described in detail. This embodiment provides a downshift user identification method, including:

[0147] The principles of S401-S404 are the same as those of S201-S204, and will not be repeated here.

[0148] S405. Determine each third user whose downgrade probability is greater than a preset probability threshold as a user whose package level is downgraded.

[0149] The probability threshold is a threshold obtained by the downshift user identification method of the first aspect of the invention.

[0150] Specifically, the probability threshold calculated in the above embodiment is compared with the downshift probability, and each third user whose downshift probability is greater than the preset probability threshold is determined as a user whose package level is downgraded.

[0151] The present application provides a method for identifying downgraded users, which uses a probability threshold to predict and determine the downgrade behavior of the user, thereby improving the accuracy of the prediction.

[0152] Figure 5 Schematic diagram of the process of downshift user identification method provided in the embodiment of the present application Figure 4 In this embodiment Figure 2 and Figure 4 Based on the above, the downshift user identification method is described in detail. The downshift user identification method provided in this embodiment includes:

[0153] The principles of S501-S505 are the same as those of S401-S405, and will not be repeated here.

[0154] S506: Obtain a second number of users who want to downshift.

[0155] The second number of downgraded users refers to the actual number of users whose package levels are downgraded among the multiple third users;

[0156] Specifically, the third user in the actual application is obtained, and the actual number of users whose package levels are reduced, that is, the second number of users.

[0157] S507: Update the probability threshold according to the downshift probabilities of the plurality of third users and the second number of downshifted users.

[0158] Specifically, according to the downshift probabilities of each of the multiple third users and the second swiping volume of the downshifted users, the probability threshold is updated through the downshift recognition model and artificial intelligence algorithm to facilitate screening out the downshifted users.

[0159] The present application provides a downgraded user identification method, which can obtain real-time user data and quantity, update the probability threshold, perform continuous optimization, and improve the adaptability of the prediction.

[0160] The embodiment of the present invention can divide the electronic device or the main control device into functional modules according to the above method example. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the embodiment of the present invention is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0161] Figure 6 A schematic diagram of the structure of the downshift user identification device provided in the embodiment of the present application Figure 1 .like Figure 6 As shown, the downshift user identification device 600 includes: a data acquisition module 610 , a model screening module 620 , a probability calculation module 630 and a judgment module 640 .

[0162] The data acquisition module 610 is used to acquire package change record data sets of multiple first users, and perform feature extraction on each package change record data set to obtain first data sets of multiple first users; wherein the first data set refers to a data set related to the package level;

[0163] A model screening module 620, configured to screen a plurality of second data sets from a plurality of first data sets by using a preset downshift recognition model;

[0164] The probability calculation module 630 is used to calculate the downgrade probability of each second data set through a preset artificial intelligence algorithm; wherein the downgrade probability refers to the probability of the user's package level being reduced;

[0165] The judgment module 640 is used to obtain a probability threshold according to the downgrade probability of each second data set and the first number of downgraded users obtained, so as to judge whether the package level of the user is downgraded according to the probability threshold; wherein the first number of downgraded users refers to the actual number of users whose package levels are downgraded among multiple first users.

[0166] In a possible design, the plurality of downshift probabilities are all within a first value range;

[0167] The determination module 640 includes:

[0168] A numerical range module, used to divide the first numerical range into multiple second numerical ranges according to multiple preset division thresholds; wherein the minimum value of each second numerical range is equal to a division threshold, and the maximum value of each second numerical range is equal to the maximum value of the first numerical range;

[0169] A user quantity module, for obtaining the cumulative number of users for each division threshold according to a plurality of downshift probabilities; wherein the cumulative number of users refers to the number of downshift probabilities within the corresponding second numerical range;

[0170] The threshold determination module is used to determine a probability threshold from multiple division thresholds according to the accumulated number of users of each division threshold and the first number of downgraded users.

[0171] In a possible design, a threshold determination module includes:

[0172] A quantity difference module, used for calculating the quantity difference between the accumulated quantity of each user and the first quantity of downshifted users;

[0173] The partitioning threshold module is used to determine the partitioning threshold with the smallest absolute value of the quantity difference as the probability threshold.

[0174] In a possible design, each second data set includes a plurality of characteristic data, and the characteristic data refers to package-level related data;

[0175] The probability calculation module 630 includes:

[0176] A weighted calculation module, used to perform weighted calculation on the plurality of characteristic data of each second data set to obtain a weighted calculation result of each second data set;

[0177] A mapping module, used to map the multiple weighted calculation results to a first numerical range through a preset mapping function to obtain a mapping result of each weighted calculation result;

[0178] The downshift probability module is used to determine the mapping result of each weighted calculation result as the downshift probability of the corresponding second data set.

[0179] In a possible design, each first data set includes: primary and secondary card identification information, card activation record information, package level information, and package value information;

[0180] The second user is any one of the plurality of first users. For the second user, the model screening module 620 includes:

[0181] The switching module is used to determine the first data set of the second user as the second data set when the downgrade recognition model determines that the previous number of the second user is switched from the main card to the secondary card of the newly opened number, the current package level of the newly opened number is lower than the original package level of the previous number, and the current package value of the second user is lower than the original package level of the second user.

[0182] In a possible design, the downshift user identification device 600 further includes:

[0183] The data cleaning module is used to clean each package change record data set separately;

[0184] The data processing module is used to standardize each package change record data set after data cleaning.

[0185] The present embodiment provides a downshift user identification device, which can execute the downshift user identification method of the above-mentioned embodiment. Its implementation principle and technical effect are similar, and will not be described in detail in the present embodiment.

[0186] Figure 7 A schematic diagram of the structure of the downshift user identification device provided in the embodiment of the present application Figure 2 .like Figure 7 As shown, the downshift user identification device includes: an acquisition probability module 710 and a user screening module 720 .

[0187] An acquisition probability module 710 is used to acquire the downshift probability of each of a plurality of third users;

[0188] The user screening module 720 is used to determine each third user whose downgrade probability is greater than a preset probability threshold as a user whose package level is downgraded; wherein the probability threshold is a threshold obtained by the downgrade user identification device of the above embodiment.

[0189] In a possible design, the downshift user identification device further includes:

[0190] An acquisition module for acquiring a second number of downgraded users, wherein the second number of downgraded users refers to the actual number of users whose package levels are downgraded among the plurality of third users;

[0191] The threshold updating module is used to update the probability threshold according to the downshift probabilities of the plurality of third users and the second number of downshifted users.

[0192] The present embodiment provides a downshift user identification device, which can execute the downshift user identification method of the above-mentioned embodiment. Its implementation principle and technical effect are similar, and will not be described in detail in the present embodiment.

[0193] In the specific implementation of the aforementioned downshift user identification method, each module can be implemented as a processor, and the processor can execute computer-executable instructions stored in the memory, so that the processor executes the aforementioned downshift user identification method.

[0194] Figure 8 This is a schematic diagram of the structure of the electronic device hardware provided in the embodiment of the present application. Figure 8 As shown, the electronic device 800 includes: at least one processor 810 and a memory 820. The electronic device 800 also includes a communication component 830. The processor 810, the memory 820 and the communication component 830 are connected via a bus 840.

[0195] In a specific implementation process, at least one processor 810 executes the computer-executable instructions stored in the memory 820, so that at least one processor 810 executes a downshift user identification method executed by the electronic device side as above.

[0196] The specific implementation process of the processor 810 can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0197] In the above embodiments, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the invention may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0198] The memory may include a high-speed RAM memory, and may also include a non-volatile storage NVM, such as at least one disk storage.

[0199] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of this application is not limited to only one bus or one type of bus.

[0200] The above-mentioned functions implemented by the electronic device and the main control device introduce the scheme provided by the embodiment of the present invention. It can be understood that in order to implement the above-mentioned functions, the electronic device or the main control device includes a hardware structure and / or software module corresponding to the execution of each function. In combination with the units and algorithm steps of each example described in the embodiment disclosed in the embodiment of the present invention, the embodiment of the present invention can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the technical solution of the embodiment of the present invention.

[0201] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the above-mentioned downshift user identification method.

[0202] The above-mentioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof.

[0203] Such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special-purpose computer.

[0204] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). Of course, the processor and the readable storage medium can also exist as discrete components in an electronic device or a main control device.

[0205] The present application also provides a computer program product, which includes: a computer program, which is stored in a readable storage medium. At least one processor of an electronic device can read the computer program from the readable storage medium. When the computer program is executed by the processor, it is used to implement the above-mentioned downshift user identification method.

[0206] Those skilled in the art can understand that all or part of the steps of the above method embodiments can be completed by hardware related to program instructions. The above program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above method embodiments are executed; and the above storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0207] So far, the technical solution of the present application has been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present application is obviously not limited to these specific embodiments, and the above embodiments are only used to illustrate the technical solution of the present application rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A downshift user identification method, characterized in that: The method comprises: Acquire package change record data sets of respective multiple first users, and perform feature extraction on each of the package change record data sets to obtain first data sets of respective multiple first users; wherein the first data sets refer to relevant data sets of package level; Filtering a plurality of second data sets from a plurality of the first data sets by using a preset downshift recognition model; The downgrade probability of each of the second data sets is calculated by a preset artificial intelligence algorithm; wherein the downgrade probability refers to the probability that the package level of the user is downgraded; A probability threshold is obtained based on the downgrade probability of each of the second data sets and the first number of downgraded users obtained, so as to determine whether the package level of the user is downgraded based on the probability threshold; wherein the first number of downgraded users refers to the actual number of users whose package levels are downgraded among the multiple first users.

2. The downshift user identification method according to claim 1, characterized in that: A plurality of the downshift probabilities are all within a first value range; The obtaining a probability threshold according to the downshift probability of each of the second data sets and the acquired first number of downshifted users includes: According to a plurality of preset division thresholds, the first numerical range is divided into a plurality of second numerical ranges; wherein the minimum value of each of the second numerical ranges is equal to one of the division thresholds, and the maximum value of each of the second numerical ranges is equal to the maximum value of the first numerical range; According to the plurality of the downshift probabilities, the accumulated number of users for each of the division thresholds is obtained by summarizing; wherein the accumulated number of users refers to the number of downshift probabilities within the corresponding second numerical range; The probability threshold is determined from the multiple division thresholds according to the accumulated number of users of each division threshold and the first number of downshifted users.

3. The downshift user identification method according to claim 2, characterized in that: The step of determining the probability threshold from the plurality of division thresholds according to the accumulated number of users of each division threshold and the first number of downgraded users comprises: Calculating the difference between the accumulated number of each user and the first number of downshifted users; The division threshold with the smallest absolute value of the quantity difference is determined as the probability threshold.

4. The downshift user identification method according to claim 2 or 3, characterized in that: Each of the second data sets includes a plurality of feature data, wherein the feature data refers to package level related data; The step of calculating the downshift probability of each of the second data sets by a preset artificial intelligence algorithm includes: Performing weighted calculation on the plurality of characteristic data of each of the second data sets to obtain a weighted calculation result of each of the second data sets; Mapping the plurality of weighted calculation results to the first numerical range through a preset mapping function to obtain a mapping result of each weighted calculation result; A mapping result of each of the weighted calculation results is determined as the downshift probability of the corresponding second data set.

5. The downshift user identification method according to claim 1, characterized in that: Each of the first data sets includes: primary and secondary card identification information, card activation record information, package level information and package value information; The second user is any one of the plurality of first users. For the second user, the plurality of second data sets are screened out from the plurality of first data sets by using a preset downshift recognition model, including: When the downgrade identification model determines that the previous number of the second user is switched from a main card to a secondary card of a newly opened number, the current package level of the newly opened number is lower than the original package level of the previous number, and the current package value of the second user is lower than the original package level of the second user, the first data set of the second user is determined as the second data set.

6. The downshift user identification method according to claim 1, characterized in that: Before extracting features from each of the package change record data sets to obtain respective first data sets of the plurality of first users, the method further includes: Performing data cleaning on each of the package change record data sets respectively; Each package change record data set after data cleaning is standardized.

7. A downshift user identification method, characterized in that: The method comprises: Obtaining respective downshift probabilities of a plurality of third users; Each of the third users whose downgrade probability is greater than a preset probability threshold is determined as a user whose package level is downgraded; wherein the probability threshold is a threshold obtained by the downgrade user identification method according to any one of claims 1 to 6.

8. The downshift user identification method according to claim 7, characterized in that: After determining each of the third users whose downgrade probability is greater than a preset probability threshold as a user with a lowered package level, the method further includes: Acquire a second number of downgraded users; wherein the second number of downgraded users refers to the actual number of users whose package levels are downgraded among the plurality of third users; The probability threshold is updated according to the downshift probabilities of the respective third users and the second number of downshifted users.

9. A downshift user identification device, characterized in that: include: A data acquisition module is used to acquire package change record data sets of multiple first users, and perform feature extraction on each package change record data set to obtain first data sets of the multiple first users; wherein the first data set refers to a data set related to the package level; A model screening module, used for screening a plurality of second data sets from a plurality of the first data sets by using a preset downshift recognition model; A probability calculation module, used to calculate the downgrade probability of each of the second data sets through a preset artificial intelligence algorithm; wherein the downgrade probability refers to the probability that the package level of the user is downgraded; A judgment module is used to obtain a probability threshold according to the downgrade probability of each of the second data sets and the first number of downgraded users obtained, so as to judge whether the package level of the user is reduced according to the probability threshold; wherein the first number of downgraded users refers to the actual number of users whose package levels are reduced among the multiple first users.

10. A downshift user identification device, characterized in that: include: An acquisition probability module, used to acquire the downshift probability of each of a plurality of third users; The user screening module is used to determine each of the third users whose downgrade probability is greater than a preset probability threshold as a user whose package level is downgraded; wherein the probability threshold is a threshold obtained by the downgrade user identification device according to claim 9.

Images