Prepaid card management method and system based on artificial intelligence

Abstract

The invention relates to the technical field of card management, in particular to a prepaid card management method and system based on artificial intelligence, and the method comprises the following steps: S1, carrying out the collection and feature analysis of the balance and consumption behavior data of a user; s2, merchant preferential information collection and feature modeling; s3, intelligently generating a user balance incentive rule; and S4, matching and optimizing the merchant preferences and the user demands. According to the invention, the possible consumption intention of the user can be predicted according to the specific consumption behavior of the user, and the incentive measure for balancing the use of the balance of the user is automatically triggered at a proper time, so that the interaction between the merchant and the user is optimized, the acceptability and the use frequency of the user to the preferences of the merchant are improved, and the consumption conversion rate of the merchant is effectively increased. And the overall user experience of the prepaid card system is further enhanced.

Description

Technical Field

[0001] The present invention relates to the field of card management technology, and in particular to a prepaid card management method and system based on artificial intelligence. Background Art

[0002] With the popularization of electronic payment and changes in consumer shopping behavior, prepaid cards (such as recharge cards, gift cards, etc.) have gradually become a common payment tool. Prepaid cards can not only be used for consumption at merchants, but also provide consumers with a certain way to pre-deposit funds, which is convenient, fast and safe. Therefore, prepaid cards are increasingly used in various scenarios, such as e-commerce, retail industry, catering and entertainment, and transportation.

[0003] However, most current prepaid card management systems have certain limitations in the intelligent matching of user usage and merchant discounts. Existing technologies mainly rely on simple basic functions such as balance inquiry, recharge and consumption records, and fail to fully utilize the consumption behavior characteristics of users to intelligently improve consumption experience and optimize capital utilization efficiency. In addition, merchant discount recommendations are usually based on fixed rules or users' historical consumption data, lacking a personalized recommendation mechanism, resulting in a low user response rate to merchant discounts. Specific problems include:

[0004] Idle balance problem: Many users have idle balances on their prepaid cards, and the balances are not fully utilized, resulting in low liquidity, especially when users have low consumption frequency or expired balances.

[0005] Lack of personalized incentive mechanism: Existing prepaid card management systems usually adopt unified incentive measures (such as fixed cash back or discounts), and fail to make dynamic adjustments based on factors such as users' consumption behavior patterns, consumption time and merchant category preferences, resulting in reduced effectiveness of incentive measures and difficulty in stimulating users' consumption motivation.

[0006] Insufficient accuracy in merchant discount recommendations: Although merchant discount information has been used to a certain extent in many prepaid card systems, most discount recommendation systems do not take into account the matching between users’ personalized needs and merchants’ discount conditions. The recommendation results are usually not accurate enough to effectively improve users’ acceptance and usage of discounts. Summary of the invention

[0007] The present invention provides a prepaid card management method and system based on artificial intelligence.

[0008] A prepaid card management method based on artificial intelligence, comprising the following steps:

[0009] S1, user balance and consumption behavior data collection and feature analysis:

[0010] Use the prepaid card management system to collect users' balance data, consumption records and frequently used merchant information in real time; perform feature analysis on users' historical consumption behaviors and generate personalized consumption behavior feature vectors for users, including balance usage patterns, consumption time preferences and merchant category preferences;

[0011] S2, merchant discount information collection and feature modeling:

[0012] Establish a merchant feature database in the prepaid card management system, collect dynamic discount information of merchants in real time through the merchant-side interface, including discount amount, discount activity time window and applicable consumption conditions, use clustering algorithm to classify merchant discount information, and generate merchant discount feature matrix;

[0013] S3, intelligent generation of user balance incentive rules:

[0014] Based on the consumption behavior feature vector of the user, dynamic balance incentive rules are constructed. The dynamic balance incentive rules include incentive priority and incentive form. The incentive form includes additional balance cashback, discount coupons or point rewards.

[0015] S4, optimization of matching between merchant discounts and user needs:

[0016] Based on the merchant discount feature matrix, it is matched with the incentive form, and preferential information compatible with the balance incentive rules is recommended first for matching and sorting merchant discount information.

[0017] Optionally, the S1 specifically includes:

[0018] S11, real-time data collection: through the user data interface in the prepaid card management system, establish a connection with the user consumption terminal (such as POS machine, online payment system) and the merchant database, and the collected data includes:

[0019] Balance data: user's current card balance and historical balance changes;

[0020] Consumption records: including consumption amount, payment method, and consumption time;

[0021] Frequently visited merchant information: Based on consumption frequency, filter the merchants and merchant categories that users frequently visit.

[0022] S12, data preprocessing: normalize the collected raw data, organize the data by user dimension and store it in a time series data format for subsequent model training;

[0023] S13, based on the collected consumption behavior data, generating a personalized consumption behavior feature vector, including:

[0024] Balance usage pattern: By analyzing the balance change trend, the volatility of the user's balance usage is extracted to reflect consumption habits;

[0025] Consumption time preference: Analyze the consumption time distribution characteristics of users and use time segments to analyze consumption probability;

[0026] Merchant category preference: Calculate the user's preference weight for merchants in each category based on the merchant category labels of the commonly used merchant information in the user's consumption records.

[0027] Optionally, the balance usage pattern is extracted by analyzing the time series of the user's balance change (such as high-frequency consumption or long-term non-use), which is reflected as the balance volatility and calculated as:

[0028] Among them, ΔB is the balance volatility, which reflects the stability of the user's use of the balance, and B u is the balance value after the u-th consumption, is the average balance value of the user during the observation period, N is the number of consumptions during the observation period, and the higher the balance volatility ΔB, the more frequent or unstable the user's balance use; the lower the balance volatility ΔB, the more likely the user is to maintain a stable balance;

[0029] The consumption time preference includes analyzing the consumption time distribution characteristics of users, and using the time segment consumption probability to analyze: Among them, p t is the consumption probability of the user in the tth time period, n t is the number of times the user consumes in the tth time period, and T is the total number of time periods (for example, a day is divided into 24-hour periods, then is the total number of consumption times of the user in all time periods of a day, j is the time index, and the distribution p t Describe the user's consumption behavior preferences at different times of the day;

[0030] The merchant category preference includes weight calculation, which is calculated by counting the consumption frequency of different merchant categories in the user's consumption records: Among them, w i is the user’s preference weight for the i-th merchant, f i is the consumption frequency of the user at the i-th merchant, q is the total number of merchant categories, f j represents the total consumption frequency of all merchant categories of the user, j index ranges from 1 to T (total number of time periods), and the preference weight w i It reflects the user's consumption tendency towards different merchant categories. The higher the value, the stronger the user's preference for merchants in that category.

[0031] Optionally, the consumption behavior feature vector integrates the features of each dimension into a user's personalized consumption behavior feature vector F, which is expressed as: Among them, ΔB is the balance usage pattern characteristic, is the merchant category preference feature, with dimension N (total number of merchant categories), It is the consumption time preference characteristic, and its dimension is T (total number of time periods).

[0032] Optionally, the merchant preferential information collection and feature modeling in S2 specifically include:

[0033] S21, establishment of merchant feature database: In the prepaid card management system, the merchant's basic information and promotional information are collected in real time through the merchant interface to establish a merchant feature database. The merchant interface is connected to the merchant system through the API to collect and store the following merchant data:

[0034] Merchant basic information: including merchant ID, merchant category, geographic location, and business hours;

[0035] Dynamic discount information: including discount amount, discount activity time window, and applicable consumption conditions;

[0036] S22, clustering algorithm for feature classification: using clustering algorithm to classify the dynamic discount information of merchants, so as to provide users with more accurate recommendations, including:

[0037] The K-means clustering algorithm is used to perform unsupervised learning on the merchant's preferential information and cluster them according to the following characteristics:

[0038] Discount amount d: The discount amount or discount ratio is used as the clustering dimension;

[0039] Promotion time window (t start ,t end ): The time period of the activity (such as holidays, weekends or weekdays) is used as the clustering dimension;

[0040] Applicable consumption conditions (c min ,c max ): The consumption category (such as catering, retail, entertainment, etc.) is used as the clustering dimension.

[0041] The clustering results generate a merchant discount feature matrix, which includes the average discount rate, applicable time period, and target consumer group characteristics for each type of discount activity, making it easier to match appropriate discount information based on user needs.

[0042] Optionally, the K-means clustering algorithm classifies merchants according to their preferential features to form a merchant preferential feature matrix, and minimizes the Euclidean distance between merchants through iterative optimization, with the goal of minimizing the cost function (loss function): Among them, K is the number of categories (cluster number) of merchant discount information, that is, how many different discount categories the merchant is divided into, N k is the number of merchants in the kth category, x i is the characteristic vector of the ith merchant, including discount amount, preferential time window, consumption conditions, etc., μ k : The cluster center of the kth category (i.e. the average value of the characteristics of merchants in this category).

[0043] The process of Kmeans clustering algorithm is as follows:

[0044] 1. Initialization: Randomly select K merchants as the initial cluster centers;

[0045] 2. Assign clusters: For each merchant, calculate its distance to the center of each cluster and assign the merchant to the nearest cluster;

[0046] 3. Update cluster center: Update the cluster center according to the average value of the characteristics of merchants in each cluster;

[0047] 4. Iteration: Repeat steps 2 and 3 until the cluster centers converge (i.e., the cluster centers no longer change);

[0048] After the cluster analysis is completed, each merchant will correspond to a discount feature vector, including the merchant’s discount, discount activity time, and consumption conditions. The merchant discount feature matrix M after clustering is: M = [m1, m2, …, m i ], where m i Represents the preferential feature vector of the ith merchant. After normalization and clustering, the feature set represents the comprehensive characteristics of the merchant's preferential activities. This feature matrix serves as the input data for the subsequent optimization of matching merchant preferential offers with user needs.

[0049] Optionally, the incentive priority in S3 is calculated by sorting according to the comprehensive score G as follows:

[0050] Among them, α1, α2, α3 are the weight coefficients of each feature, and the incentive priority is obtained by sorting according to the comprehensive score G of each prepaid card user.

[0051] Optionally, the incentive form is formulated according to the consumption behavior feature vector F:

[0052] Additional balance cashback: When the user's balance is used less frequently and has a large balance, the incentive form is balance cashback (for example, returning a certain percentage of the balance);

[0053] Discount coupons: When users frequently use prepaid cards for small purchases, the incentive is in the form of discount coupons (for example, discount coupons can be used when spending at specific merchants);

[0054] Points rewards: When users have diverse consumption categories or long-term consumption habits, the incentive form is points rewards (for example, points can be obtained for every certain amount of consumption, and points can be used to redeem goods or services).

[0055] Optionally, the optimization of matching merchant discounts with user needs in S4 specifically includes:

[0056] Matching of additional balance cashback: Select the matching cashback incentive promotion information, considering whether the merchant has discount activities or promotions that are compatible with the balance cashback. For example, if the merchant's promotion offers preferential conditions such as "cashback for purchases over a certain amount" or "cashback for spending over a certain amount", it will match the balance cashback incentive.

[0057] Matching of discount coupons: Choose promotional activities that match discount coupons, and consider whether the merchant offers "specific discounts", "discount promotions" or "limited-time discounts". Discount coupon incentives are suitable for scenarios where users can get price discounts immediately when consuming.

[0058] Matching of points rewards: Choose promotional activities that match point rewards, and consider whether the merchant offers "points exchange" or "consumption accumulation points" activities. These merchant offers are usually related to the user's continuous consumption habits.

[0059] Based on the matching results of the merchant discount feature matrix and the incentive form, a merchant discount matching list is generated, and the merchant discounts are sorted according to the matching degree.

[0060] An artificial intelligence-based prepaid card management system, used to implement the above-mentioned artificial intelligence-based prepaid card management method, includes the following modules:

[0061] Data collection module, used to collect users' balance data, consumption records and frequently used merchant information in real time;

[0062] The user feature analysis module analyzes the user's historical consumption behavior and generates a personalized consumption behavior feature vector. The features include balance usage pattern, consumption time preference and merchant category preference.

[0063] The merchant discount information collection module collects the dynamic discount information of merchants in real time through the merchant terminal interface, including discount amount, discount activity time window and applicable consumption conditions, and stores the collected merchant discount information in the merchant feature database;

[0064] The merchant discount feature modeling module is used to classify the merchant discount information using a clustering algorithm and generate a merchant discount feature matrix;

[0065] A dynamic balance incentive rule generation module, which intelligently generates dynamic balance incentive rules based on the user consumption behavior feature vector. The incentive rules include incentive priority and incentive form;

[0066] The discount matching and recommendation module matches the merchant discount feature matrix with the user's dynamic balance incentive rules, gives priority to recommending discount information that is compatible with the balance incentive rules, and sorts the merchant discount information according to the recommendation results.

[0067] Beneficial effects of the present invention:

[0068] The present invention, by combining the user's personalized consumption behavior feature vector, identifies the user's consumption tendency and demand, realizes the intelligent generation and optimization of dynamic balance incentive rules, and through in-depth analysis of the user's balance usage pattern, consumption time preference and merchant category preference, can predict the user's possible consumption intention according to the user's specific consumption behavior, and automatically trigger incentive measures to balance the user's balance usage at an appropriate time. In particular, when the balance is about to be idle or the consumption frequency decreases, dynamic incentives (such as cash back, points or discounts) can effectively stimulate user consumption, avoid long-term unused balances, and improve the liquidity of prepaid cards and the user's consumption activity.

[0069] The present invention improves the effect of merchant discounts by establishing a merchant discount feature matrix and using intelligent recommendation to accurately match discount information with incentive measures, combines incentive forms with merchant discounts, and gives priority to recommending discount information that is compatible with dynamic balance incentive rules (for example, recommending cash back discounts to users with unused balances and recommending discount coupons to regular consumer users), thereby optimizing the interaction between merchants and users, improving users' acceptance of merchant discounts and usage frequency, effectively increasing merchants' consumption conversion rates, and further enhancing the overall user experience of the prepaid card system. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0071] Figure 1 A schematic diagram of a prepaid card management method according to an embodiment of the present invention;

[0072] Figure 2 Schematic diagram of a prepaid card management system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0073] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0074] It should be noted that the references to "one embodiment", "an embodiment", "an exemplary embodiment", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0075] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0076] like Figure 1 As shown, a prepaid card management method based on artificial intelligence includes the following steps:

[0077] S1, user balance and consumption behavior data collection and feature analysis:

[0078] Use the prepaid card management system to collect users' balance data, consumption records and frequently used merchant information in real time; perform feature analysis on users' historical consumption behaviors and generate personalized consumption behavior feature vectors for users, including balance usage patterns, consumption time preferences and merchant category preferences;

[0079] S2, merchant discount information collection and feature modeling:

[0080] Establish a merchant feature database in the prepaid card management system, collect dynamic discount information of merchants in real time through the merchant-side interface, including discount amount, discount activity time window and applicable consumption conditions, use clustering algorithm to classify merchant discount information, and generate merchant discount feature matrix;

[0081] S3, intelligent generation of user balance incentive rules:

[0082] Based on the consumption behavior feature vector of the user, dynamic balance incentive rules are constructed. The dynamic balance incentive rules include incentive priority and incentive form. The incentive form includes additional balance cashback, discount coupons or point rewards.

[0083] S4, optimization of matching between merchant discounts and user needs:

[0084] Based on the merchant discount feature matrix, it is matched with the incentive form, and preferential information compatible with the balance incentive rules is recommended first for matching and sorting merchant discount information.

[0085] S1 specifically includes:

[0086] S11, real-time data collection: through the user data interface in the prepaid card management system, establish a connection with the user consumption terminal (such as POS machine, online payment system) and the merchant database, and the collected data includes:

[0087] Balance data: user's current card balance and historical balance changes;

[0088] Consumption records: including consumption amount, payment method, and consumption time;

[0089] Frequently visited merchant information: Based on consumption frequency, filter the merchants and merchant categories that users frequently visit.

[0090] S12, data preprocessing: normalize the collected raw data, organize the data by user dimension and store it in a time series data format for subsequent model training;

[0091] The normalization process specifically includes: in,

[0092] x is the original data value, x′ is the normalized data value, and x min is the minimum value of the feature in the data set, x max It is the maximum value of the feature in the data set. Normalization ensures that the data value is in the range of [0,1] and eliminates the impact of dimensional differences on subsequent models.

[0093] S13, based on the collected consumption behavior data, generating a personalized consumption behavior feature vector, including:

[0094] Balance usage pattern: By analyzing the balance change trend, the volatility of the user's balance usage is extracted to reflect consumption habits;

[0095] Consumption time preference: Analyze the consumption time distribution characteristics of users and use time segments to analyze consumption probability;

[0096] Merchant category preference: Calculate the user's preference weight for merchants in each category based on the merchant category labels of the commonly used merchant information in the user's consumption records.

[0097] Balance usage pattern By analyzing the time series of user balance changes, the usage pattern (such as high-frequency consumption or long-term non-use) is extracted and reflected as balance volatility, which is calculated as:

[0098] Among them, ΔB is the balance volatility, which reflects the stability of the user's use of the balance, and B u is the balance value after the u-th consumption, is the average balance value of the user during the observation period, N is the number of consumptions during the observation period, and the higher the balance volatility ΔB, the more frequent or unstable the user's balance use; the lower the balance volatility ΔB, the more likely the user is to maintain a stable balance;

[0099] Consumption time preference includes analyzing the consumption time distribution characteristics of users, using the time segment consumption probability to analyze: Among them, p t is the consumption probability of the user in the tth time period, n t is the number of times the user consumes in the tth time period, T is the total number of time periods (for example, if a day is divided into 24 hours, then T = 24), is the total number of consumption times of the user in all time periods of a day, j is the time index, and the distribution p t Describe the user's consumption behavior preferences at different times of the day;

[0100] Merchant category preference includes weight calculation, which is calculated by counting the consumption frequency of different merchant categories in the user's consumption records: Among them, w i is the user’s preference weight for the i-th merchant, f i is the consumption frequency of the user at the i-th merchant, q is the total number of merchant categories, represents the total consumption frequency of all merchant categories of the user, j index ranges from 1 to T (total number of time periods), and the preference weight w i It reflects the user's consumption tendency towards different merchant categories. The higher the value, the stronger the user's preference for merchants in that category.

[0101] The consumption behavior feature vector integrates the features of each dimension into the user's personalized consumption behavior feature vector F, which is expressed as: Among them, ΔB is the balance usage pattern characteristic, is the merchant category preference feature, with dimension N (total number of merchant categories), It is the consumption time preference characteristic, and its dimension is T (total number of time periods).

[0102] The merchant discount information collection and feature modeling in S2 specifically include:

[0103] S21, establishment of merchant feature database: In the prepaid card management system, the merchant's basic information and promotional information are collected in real time through the merchant interface to establish a merchant feature database. The merchant interface is connected to the merchant system through the API to collect and store the following merchant data:

[0104] Merchant basic information: including merchant ID, merchant category, geographic location, and business hours;

[0105] Dynamic discount information: including discount amount, discount activity time window, and applicable consumption conditions. Specifically:

[0106] Discount amount: record the discount amount of the promotion (for example, 20 yuan off for purchases over 100 yuan, or 20% off);

[0107] Promotion time window: record the start and end time of promotions;

[0108] Applicable consumption conditions: including consumption amount, specific product or service categories, user type and other restrictions.

[0109] Using the merchant-side interface, the prepaid card management system pulls the latest discount information from the merchant system regularly or in real time to ensure that the discount information in the database is up to date. To avoid data conflicts, the system uses a timestamp synchronization mechanism to ensure that relevant data is not lost or missed when discount activities are updated.

[0110] S22, clustering algorithm for feature classification: using clustering algorithm to classify the dynamic discount information of merchants, so as to provide users with more accurate recommendations, including:

[0111] The K-means clustering algorithm is used to perform unsupervised learning on the merchant's preferential information and cluster them according to the following characteristics:

[0112] Discount amount d: The discount amount or discount ratio is used as the clustering dimension;

[0113] Promotion time window (t start ,t end ): The time period of the activity (such as holidays, weekends or weekdays) is used as the clustering dimension;

[0114] Applicable consumption conditions (c min ,c max ): The consumption category (such as catering, retail, entertainment, etc.) is used as the clustering dimension.

[0115] The clustering results generate a merchant discount feature matrix, which includes the average discount rate, applicable time period, and target consumer group characteristics for each type of discount activity, making it easier to match appropriate discount information based on user needs.

[0116] The K-means clustering algorithm classifies merchants according to their preferential features to form a merchant preferential feature matrix. Through iterative optimization, the Euclidean distance between merchants is minimized. The goal is to minimize the cost function (loss function): Among them, K is the number of categories (cluster number) of merchant discount information, that is, how many different discount categories the merchant is divided into, N k is the number of merchants in the kth category, x i is the characteristic vector of the ith merchant, including discount amount, preferential time window, consumption conditions, etc., μ k : The cluster center of the kth category (i.e. the average value of the characteristics of merchants in this category).

[0117] The process of Kmeans clustering algorithm is as follows:

[0118] 1. Initialization: Randomly select K merchants as the initial cluster centers;

[0119] 2. Assign clusters: For each merchant, calculate its distance to the center of each cluster and assign the merchant to the nearest cluster;

[0120] 3. Update cluster center: Update the cluster center according to the average value of the characteristics of merchants in each cluster;

[0121] 4. Iteration: Repeat steps 2 and 3 until the cluster centers converge (i.e., the cluster centers no longer change);

[0122] After the cluster analysis is completed, each merchant will correspond to a discount feature vector, including the merchant’s discount, discount activity time, and consumption conditions. The merchant discount feature matrix M after clustering is: M = [m1, m2, …, m i ], where m i Represents the preferential feature vector of the i-th merchant. The feature set after normalization and clustering represents the comprehensive characteristics of the merchant's preferential activities. The feature matrix serves as the input data for the subsequent optimization of matching merchant discounts with user needs.

[0123] m i Includes the following properties:

[0124] Discount amount: such as a percentage or a fixed amount of the discount.

[0125] Promotion time window: such as the time period during which the promotion is valid (for example, weekends, holidays, etc.).

[0126] Applicable consumption conditions: such as whether there is a minimum consumption limit, whether it can be used in conjunction with other discounts, etc. The incentive priority in S3 is calculated based on the comprehensive score G:

[0127] Among them, α1, α2, and α3 are weight coefficients of each feature, which combine the balance usage pattern, merchant preference, and time preference to generate an incentive priority suitable for the user. The incentive priority is obtained by sorting according to the comprehensive score G of each prepaid card user (since the incentive pool quota is limited, the higher the incentive priority, the earlier the incentive is issued. When the incentive pool is used up, the incentive is closed. Different incentive quotas can also be set according to the incentive priority).

[0128] The incentive form is formulated according to the consumption behavior feature vector F:

[0129] Additional balance cashback: When the user's balance is used less frequently and has a large balance, the incentive form is balance cashback (for example, returning a certain percentage of the balance);

[0130] Discount coupons: When users frequently use prepaid cards for small purchases, the incentive is in the form of discount coupons (for example, discount coupons can be used when spending at specific merchants);

[0131] Points rewards: When users have diverse consumption categories or long-term consumption habits, the incentive form is points rewards (for example, points can be obtained for every certain amount of consumption, and points can be used to redeem goods or services).

[0132] The optimization of matching merchant discounts with user needs in S4 specifically includes:

[0133] Matching of additional balance cashback: Select the matching cashback incentive promotion information, considering whether the merchant has discount activities or promotions that are compatible with the balance cashback. For example, if the merchant's promotion offers preferential conditions such as "cashback for purchases over a certain amount" or "cashback for spending over a certain amount", it will match the balance cashback incentive.

[0134] Priority matching: The system prioritizes merchant discounts that meet the "balance cashback" format, such as discounts on purchases above a certain amount, cashback, etc., to ensure compatibility with the user's balance usage scenarios.

[0135] Matching of discount coupons: Choose promotional activities that match discount coupons, and consider whether the merchant offers "specific discounts", "discount promotions" or "limited-time discounts". Discount coupon incentives are suitable for scenarios where users can get price discounts immediately when consuming.

[0136] Priority matching: The system gives priority to recommending merchant discounts with clear discount conditions, such as "discount coupons" and "full-discount coupons", etc., to increase the attractiveness of user consumption.

[0137] Matching of points rewards: Choose promotional activities that match point rewards, and consider whether the merchant offers "points exchange" or "consumption accumulation points" activities. These merchant offers are usually related to the user's continuous consumption habits.

[0138] Priority matching: The system gives priority to recommending merchant offers that provide point rewards, such as "consumption points" or "free points for accumulated consumption", etc., to promote users' long-term consumption and loyalty.

[0139] Based on the matching results of the merchant discount feature matrix and the incentive form, a merchant discount matching list is generated, and the merchant discounts are sorted according to the matching degree. The sorting basis may include:

[0140] Offers with high matching degree are recommended first: If the merchant offer is compatible with the current incentive form, it will be given higher priority.

[0141] Match incentive forms with user needs: Based on the user's consumption behavior analysis and dynamic balance incentive rules, merchant discounts that match the user's current needs (such as idle balance, consumption time, etc.) are recommended first.

[0142] Finally, the merchant discount matching list generated by the system is sorted by priority and provided to users to ensure that users can obtain discount information that best matches their needs.

[0143] like Figure 2 As shown, an artificial intelligence-based prepaid card management system is used to implement the above-mentioned artificial intelligence-based prepaid card management method, including the following modules:

[0144] Data collection module, used to collect users' balance data, consumption records and frequently used merchant information in real time;

[0145] The user feature analysis module analyzes the user's historical consumption behavior and generates a personalized consumption behavior feature vector. The features include balance usage pattern, consumption time preference and merchant category preference.

[0146] The merchant discount information collection module collects the dynamic discount information of merchants in real time through the merchant terminal interface, including discount amount, discount activity time window and applicable consumption conditions, and stores the collected merchant discount information in the merchant feature database;

[0147] The merchant discount feature modeling module is used to classify the merchant discount information using a clustering algorithm and generate a merchant discount feature matrix;

[0148] A dynamic balance incentive rule generation module, which intelligently generates dynamic balance incentive rules based on the user consumption behavior feature vector. The incentive rules include incentive priority and incentive form;

[0149] The discount matching and recommendation module matches the merchant discount feature matrix with the user's dynamic balance incentive rules, gives priority to recommending discount information that is compatible with the balance incentive rules, and sorts the merchant discount information according to the recommendation results.

[0150] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0151] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A prepaid card management method based on artificial intelligence, characterized in that: The following steps are involved: S1, user balance and consumption behavior data collection and feature analysis: Use the prepaid card management system to collect users' balance data, consumption records and frequently used merchant information in real time; perform feature analysis on users' historical consumption behaviors and generate personalized consumption behavior feature vectors for users, including balance usage patterns, consumption time preferences and merchant category preferences; S2, merchant discount information collection and feature modeling: Establish a merchant feature database in the prepaid card management system, collect dynamic discount information of merchants in real time through the merchant-side interface, including discount amount, discount activity time window and applicable consumption conditions, use clustering algorithm to classify merchant discount information, and generate merchant discount feature matrix; S3, intelligent generation of user balance incentive rules: Based on the consumption behavior feature vector of the user, dynamic balance incentive rules are constructed. The dynamic balance incentive rules include incentive priority and incentive form. The incentive form includes additional balance cashback, discount coupons or point rewards. S4, optimization of matching between merchant discounts and user needs: Based on the merchant discount feature matrix, it is matched with the incentive form, and preferential information compatible with the balance incentive rules is recommended first for matching and sorting merchant discount information.

2. The artificial intelligence-based prepaid card management method according to claim 1, characterized in that: The S1 specifically includes: S11, real-time data collection: through the user data interface in the prepaid card management system, establish a connection with the user consumption terminal and the merchant database, and the collected data includes: Balance data: user's current card balance and historical balance changes; Consumption records: including consumption amount, payment method, and consumption time; Frequently used merchant information: Based on consumption frequency, filter the merchants and merchant categories that users frequently visit; S12, data preprocessing: normalize the collected raw data, organize the data by user dimension and store it in a time series data format; S13, based on the collected consumption behavior data, generating a personalized consumption behavior feature vector, including: Balance usage pattern: By analyzing the balance change trend, the volatility of the user's balance usage is extracted; Consumption time preference: Analyze the consumption time distribution characteristics of users and use time segments to analyze consumption probability; Merchant category preference: Calculate the user's preference weight for merchants in each category based on the merchant category labels of the commonly used merchant information in the user's consumption records.

3. The artificial intelligence-based prepaid card management method according to claim 2, characterized in that: The balance usage pattern is extracted by analyzing the time series of the user's balance changes, which is reflected in the balance volatility and calculated as: Among them, ΔB is the balance volatility, which reflects the stability of the user's use of the balance, and B u is the balance value after the u-th consumption, is the average balance value of the user during the observation period, and N is the number of consumption times during the observation period; The consumption time preference includes analyzing the consumption time distribution characteristics of users, and using the time segment consumption probability to analyze: Among them, p t is the consumption probability of the user in the tth time period, n t is the number of times the user consumes in the tth time period, T is the total number of time periods, is the total number of consumption times of the user in all time periods of a day, and j is the time index; The merchant category preference includes weight calculation, which is calculated by counting the consumption frequency of different merchant categories in the user's consumption records: Among them, w i is the user’s preference weight for the i-th merchant, f i is the consumption frequency of the user at the i-th merchant, q is the total number of merchant categories, f j Indicates the total consumption frequency of the user in all merchant categories.

4. The artificial intelligence-based prepaid card management method according to claim 3, characterized in that: The consumption behavior feature vector is expressed as: Among them, ΔB is the balance usage pattern characteristic, is the merchant category preference feature, It is the consumption time preference characteristic.

5. The artificial intelligence-based prepaid card management method according to claim 1, characterized in that: The merchant preferential information collection and feature modeling in S2 specifically include: S21, establishment of merchant feature database: In the prepaid card management system, the merchant's basic information and preferential activity information are collected in real time through the merchant terminal interface, and a merchant feature database is established to collect and store the following merchant data: Merchant basic information: including merchant ID, merchant category, geographic location, and business hours; Dynamic discount information: including discount amount, discount activity time window, and applicable consumption conditions; S22, clustering algorithm for feature classification: using clustering algorithm to classify the dynamic discount information of merchants, specifically including: The K-means clustering algorithm is used to perform unsupervised learning on the merchant's preferential information and cluster them according to the following characteristics: Discount amount d: The discount amount or discount ratio is used as the clustering dimension; Promotion time window (t start ,t end ): The activity time period is used as the clustering dimension; Applicable consumption conditions (c min ,c max ): Taking consumption category as clustering dimension; The clustering results generate a merchant discount feature matrix, which includes the average discount rate, applicable time period, and target consumer group characteristics for each type of discount activity.

6. The artificial intelligence-based prepaid card management method according to claim 5, characterized in that: The K-means clustering algorithm classifies merchants according to their preferential features to form a merchant preferential feature matrix. Through iterative optimization, the Euclidean distance between merchants is minimized. The goal is to minimize the cost function: Among them, K is the number of categories of merchant discount information, that is, how many different discount categories the merchant is divided into, N k is the number of merchants in the kth category, x i is the feature vector of the ith merchant, μ k : The cluster center of the kth class; After the cluster analysis is completed, each merchant will correspond to a discount feature vector, including the merchant’s discount, discount activity time, and consumption conditions. The merchant discount feature matrix M after clustering is: M = [m1, m2, …, m i ], where m i Represents the preferential feature vector of the i-th merchant.

7. The artificial intelligence-based prepaid card management method according to claim 4, characterized in that: The incentive priority in S3 is calculated by sorting according to the comprehensive score G: Among them, α1, α2, and α3 are the weight coefficients of each feature, and the incentive priority is obtained by sorting according to the comprehensive score G of each prepaid card user.

8. The artificial intelligence-based prepaid card management method according to claim 7, characterized in that: The incentive form is formulated according to the consumption behavior feature vector F: Additional balance cashback: When the user's balance is used less frequently and has a large balance, the incentive form is balance cashback; Discount coupons: When users frequently use prepaid cards for small purchases, the incentive is in the form of discount coupons; Points rewards: When users have diverse consumption categories or long-term consumption habits, the incentive form is points rewards.

9. The artificial intelligence-based prepaid card management method according to claim 1, characterized in that: The optimization of matching merchant discounts with user needs in S4 specifically includes: Matching of additional balance cashback: Select the matching cashback incentive offer information, considering whether the merchant has discount activities or promotions that are compatible with the balance cashback; Matching of discount coupons: Select the promotional activities that match the discount coupons, considering whether the merchant offers "specific discounts", "discount promotions" or "limited time discounts" activities; Matching of point rewards: Choose promotional activities that match point rewards, and consider whether the merchant offers "points exchange" or "points accumulation through consumption" activities; Based on the matching results of the merchant discount feature matrix and the incentive form, a merchant discount matching list is generated, and the merchant discounts are sorted according to the matching degree.

10. An artificial intelligence-based prepaid card management system, used to implement an artificial intelligence-based prepaid card management method as claimed in any one of claims 1 to 9, characterized in that: Includes the following modules: Data collection module, used to collect users' balance data, consumption records and frequently used merchant information in real time; The user feature analysis module analyzes the user's historical consumption behavior and generates a personalized consumption behavior feature vector. The features include balance usage pattern, consumption time preference and merchant category preference. The merchant discount information collection module collects the dynamic discount information of merchants in real time through the merchant terminal interface, including discount amount, discount activity time window and applicable consumption conditions, and stores the collected merchant discount information in the merchant feature database; The merchant discount feature modeling module is used to classify the merchant discount information using a clustering algorithm and generate a merchant discount feature matrix; A dynamic balance incentive rule generation module, which intelligently generates dynamic balance incentive rules based on the user consumption behavior feature vector. The incentive rules include incentive priority and incentive form; The discount matching and recommendation module matches the merchant discount feature matrix with the user's dynamic balance incentive rules, gives priority to recommending discount information that is compatible with the balance incentive rules, and sorts the merchant discount information according to the recommendation results.

Images