Hierarchical encryption storage system, method and medium for virtual card

Abstract

The application discloses a hierarchical encryption storage system, method and medium of virtual coupons, relates to the field of trusted storage, and comprises a coupon sensitivity evaluation module, a use record acquisition module, a user cancellation probability prediction module and an encryption storage module. The coupon sensitivity evaluation module is used for evaluating the sensitivity of virtual coupons according to coupon attribute characteristics. The use record acquisition module is used for acquiring current use records of virtual coupons and historical use records of similar virtual coupons. The user cancellation probability prediction module is used for predicting the cancellation probability of users in a preset time zone according to the use behavior characteristics of the users, the current use records and the historical use records. The encryption storage module is used for analyzing and determining the encryption strength of the coupons according to the coupon sensitivity and the predicted user cancellation probability, and performing encryption storage on the virtual coupons in the preset time zone. The technical problem that the encryption storage of existing virtual coupons is difficult to balance storage security and cancellation efficiency is solved, and the technical effect that the cancellation efficiency is improved while the storage security of virtual coupons is ensured is achieved.

Description

Technical Field

[0001] The present application relates to the field of trusted storage, and in particular to a hierarchical encrypted storage system, method, and medium for virtual cards and coupons. Background Art

[0002] As an important tool for modern commercial marketing, the storage security and ease of use of virtual cards and coupons are directly related to the interests and experience of merchants and users. It is crucial to ensure the safe storage and efficient redemption of virtual cards and coupons in different scenarios.

[0003] Currently, the primary approach to addressing virtual card storage security is to use a unified encryption strategy, applying the same encryption strength to all virtual cards for storage. This approach fails to consider the varying sensitivity of different virtual cards and user behavior. This leads to over-encryption of low-sensitivity cards, wasting system resources, while high-sensitivity cards may face security risks due to insufficient encryption strength. Furthermore, a unified encryption strategy struggles to accommodate the varying redemption probabilities among different users, creating a difficult balance between ensuring security and improving redemption efficiency.

[0004] In the current related technologies, the encrypted storage of virtual cards and coupons has the technical problem of being difficult to balance storage security and verification efficiency. Summary of the Invention

[0005] This application provides a hierarchical encrypted storage system, method, and medium for virtual cards and coupons. It uses technical means to evaluate the sensitivity of virtual cards and coupons and predict the probability of user redemption based on the card attribute characteristics, current and historical usage records of the virtual cards and coupons, and then analyze and determine the encryption strength of different cards and coupons, and perform differentiated encrypted storage. This solves the technical problem of the existing encrypted storage of virtual cards and coupons that is difficult to balance storage security and redemption efficiency, and achieves the technical effect of improving redemption efficiency while ensuring the storage security of virtual cards and coupons.

[0006] The present application provides a hierarchical encryption storage system for virtual cards and coupons, including: a card and coupon sensitivity evaluation module, which is used to evaluate the card and coupon sensitivity based on the card and coupon attribute characteristics of the virtual card and coupon, and output the card and coupon sensitivity; a usage record acquisition module, which is used to obtain the current usage record of the virtual card and coupon, as well as the historical usage records of similar virtual cards and coupons; a user cancellation probability prediction module, which is used to predict the user cancellation probability within a preset time zone based on the user's card and coupon usage behavior characteristics, the current usage record and the historical usage record, and output the predicted user cancellation probability; an encryption storage module, which is used to determine the card and coupon encryption strength based on the card and coupon sensitivity and the predicted user cancellation probability, and encrypt and store the virtual card and coupon within the preset time zone.

[0007] In a possible implementation, the coupon sensitivity evaluation module comprises: a coupon attribute feature reading unit, configured to read a coupon attribute feature of a virtual coupon, wherein the coupon attribute feature at least comprises a coupon type, a commodity value, a remaining valid time length, and a cancellation complexity; a sensitivity evaluation plug-in acquisition unit, configured to train a feedforward neural network using a sample coupon attribute feature set and a sample coupon sensitivity set to convergence, and obtain a sensitivity evaluation plug-in; and a coupon sensitivity evaluation unit, configured to input the coupon attribute feature into the sensitivity evaluation plug-in for evaluation, and output a coupon sensitivity.

[0008] In a possible implementation, the use record acquisition module comprises: a high-frequency information retrieval unit, configured to perform high-frequency information retrieval in a historical coupon operation database with the coupon attribute feature as a comparison constraint, and determine a plurality of same-type virtual coupons; and a historical use record acquisition unit, configured to acquire historical use records of the plurality of same-type virtual coupons.

[0009] In a possible implementation, the high-frequency information retrieval unit comprises: a first feature setting unit, configured to randomly select any feature from the coupon type, the commodity value, the remaining valid time length, and the cancellation complexity as a first feature without replacement; a similarity comparison unit, configured to perform similarity comparison in the historical coupon operation database with the first feature as a comparison constraint, and obtain a plurality of first same-type virtual coupons that satisfy a preset similarity threshold; a traversal unit, configured to continue to randomly select features for similarity comparison and screening without replacement based on the plurality of first same-type virtual coupons, until the features are traversed, and obtain a plurality of same-type virtual coupons; and a similarity comprehensive calculation unit, configured to perform similarity comprehensive calculation on the plurality of same-type virtual coupons respectively, determine a plurality of total similarities, and identify the plurality of same-type virtual coupons.

[0010] In a possible implementation, the user cancellation probability prediction module comprises: a cancellation proportion acquisition unit, configured to acquire cancellation proportions of a plurality of same-type virtual coupons in a preset historical time zone according to a coupon use behavior feature of a user, set as own sample cancellation probabilities, and obtain a plurality of own sample cancellation probabilities; a first predicted cancellation probability output unit, configured to configure a credible weight proportion according to the plurality of total similarities, and perform weighted fusion on the plurality of own sample cancellation probabilities to output a first predicted cancellation probability; a second predicted cancellation probability output unit, configured to perform user cancellation probability prediction according to the current use record to output a second predicted cancellation probability; a third predicted cancellation probability output unit, configured to perform user cancellation probability prediction according to the historical use record to output a third predicted cancellation probability; and an evaluation and determination unit, configured to evaluate and determine a predicted user cancellation probability based on the first predicted cancellation probability, the second predicted cancellation probability, and the third predicted cancellation probability.

[0011] In a possible implementation, the second predicted cancellation probability output unit comprises: a second initial predicted cancellation probability setting unit configured to statistically determine a cancellation proportion of the virtual card within a preset historical time zone according to the current use record, and set the cancellation proportion as a second initial predicted cancellation probability; a final cancellation proportion determination unit configured to collect a plurality of final cancellation proportion means of a plurality of similar virtual cards based on a historical card operation database, and determine a final cancellation proportion by weighting according to the credible weight proportion; a unit standard cancellation progress setting unit configured to obtain a card validity duration of the virtual card, set a ratio of the final cancellation proportion to the card validity duration as a unit standard cancellation progress; a real-time standard cancellation progress setting unit configured to obtain a cancelled proportion of the virtual card according to the current use record, obtain a predicted uncancellation proportion by subtracting the cancelled proportion from the final cancellation proportion, and set a ratio of the predicted uncancellation proportion to the remaining validity duration as a real-time standard cancellation progress; and a correction unit configured to set a ratio of the real-time standard cancellation progress to the unit standard cancellation progress as a correction coefficient, correct the second initial predicted cancellation probability, and output the second predicted cancellation probability.

[0012] In a possible implementation, the third predicted cancellation probability output unit comprises: a sample cancellation probability acquisition unit configured to statistically determine a cancellation proportion of the virtual card within a preset historical time zone according to the historical use record, and acquire a plurality of sample cancellation probabilities; and a weighted fusion unit configured to perform weighted fusion on the plurality of sample cancellation probabilities according to the credible weight proportion, and output the third predicted cancellation probability.

[0013] In a possible implementation, the encrypted storage module comprises: an initial card encryption strength matching unit configured to determine an initial card encryption strength according to the card sensitivity; a strength adjustment coefficient setting unit configured to set a ratio of the predicted user cancellation probability to a unit standard cancellation progress of the virtual card as a strength adjustment coefficient; and an encryption strength optimization unit configured to optimize the initial card encryption strength according to the strength adjustment coefficient, and output the card encryption strength.

[0014] The application further provides a hierarchical encryption storage method of a virtual card, comprising: performing card sensitivity evaluation according to card attribute characteristics of the virtual card, and outputting a card sensitivity; obtaining a current use record of the virtual card and a historical use record of a similar virtual card; predicting a user cancellation probability within a preset time zone according to card use behavior characteristics of a user, the current use record and the historical use record, and outputting a predicted user cancellation probability; and analyzing and determining a card encryption strength according to the card sensitivity and the predicted user cancellation probability, and performing encrypted storage on the virtual card within the preset time zone.

[0015] The application also provides a computer readable storage medium, comprising: a computer program stored thereon, which is executed by a processor to implement the hierarchical encryption storage method of virtual coupons.

[0016] The hierarchical encryption storage system, method and medium of virtual coupons provided in the application are used for evaluating the coupon sensitivity according to the coupon attribute characteristics of the virtual coupons, outputting the coupon sensitivity, acquiring the current use record of the virtual coupons and the historical use record of the same type of virtual coupons by using the use record acquisition module, predicting the user check-off probability in a preset time zone according to the coupon use behavior characteristics of the user, the current use record and the historical use record by using the user check-off probability prediction module, outputting the predicted user check-off probability, and analyzing and determining the coupon encryption strength according to the coupon sensitivity and the predicted user check-off probability by using the encryption storage module, so that the virtual coupons in the preset time zone are encrypted and stored. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings of the embodiments of the application will be briefly introduced below. The flowcharts are used to illustrate the operations performed by the system according to the embodiments of the application in the present application. It should be understood that the foregoing or the following operations are not necessarily executed in sequence. On the contrary, according to the needs, various steps can be processed in reverse order or at the same time. Meanwhile, other operations can be added to these processes, or one or more steps of operations can be removed from these processes.

[0018] Figure 1 The structural schematic diagram of the hierarchical encryption storage system of virtual coupons provided for the embodiments of the application.

[0019] Figure 2 The flowchart of the hierarchical encryption storage method of virtual coupons provided for the embodiments of the application.

[0020] Label explanation: coupon sensitivity evaluation module 10, use record acquisition module 20, user check-off probability prediction module 30, encryption storage module 40. DETAILED DESCRIPTION

[0021] The above description is only a summary of the technical solutions of the application. In order to more clearly understand the technical means of the application, the embodiments of the application can be implemented according to the content of the description, and in order to make the above and other purposes, characteristics and advantages of the application more obvious and easy to understand, the following specific embodiments of the application are described.

[0022] In order to make the purpose, technical solutions and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limiting this application. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0023] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict, and the terms “first\second” involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. The terms “including” and “having” and any variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or that are inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. The terms used herein are for the purpose of describing the embodiments of this application only.

[0024] The embodiment of the present application provides a hierarchical encryption storage system for virtual cards and coupons, such as Figure 1 As shown, the system includes:

[0025] The card / voucher sensitivity evaluation module 10 is configured to evaluate the card / voucher sensitivity based on the card / voucher attribute characteristics of the virtual card / voucher and output the card / voucher sensitivity.

[0026] Specifically, natural language processing (NLP) technology is used to extract attribute features of virtual vouchers, such as voucher type (e.g., coupon, gift card, membership card), amount range, validity period, and scope of use. A pre-trained machine learning model (e.g., decision tree, support vector machine, or neural network) is used to perform a sensitivity assessment on the extracted features. The model outputs a sensitivity level (e.g., low, medium, high) for the voucher based on preset rules or training data. The model output is calibrated based on business rules (e.g., vouchers with an amount exceeding 1,000 yuan are rated high sensitivity, while vouchers with an expiration date of less than one month are rated medium sensitivity).

[0027] For example, for a shopping coupon with an amount of 500 yuan and a validity period of 3 months, the NLP module extracts its type as "coupon", the amount as "500 yuan", and the validity period as "3 months". The model judges that the sensitivity of the coupon is "medium" according to the preset rules, because the amount is between 100-1000 yuan and the validity period is between 3-6 months. If the business rules stipulate that the sensitivity of the coupon with an amount exceeding 500 yuan is increased by one level, the final sensitivity of the coupon is adjusted to "high".

[0028] In one possible implementation, the coupon sensitivity evaluation module 10 includes: a coupon attribute feature reading unit for reading the coupon attribute features of a virtual coupon, wherein the coupon attribute features at least include the coupon type, the product value, the remaining valid time length, and the cancellation complexity; a sensitivity evaluation plug-in acquisition unit for training a feedforward neural network to convergence using a sample coupon attribute feature set and a sample coupon sensitivity set, and harvesting a sensitivity evaluation plug-in; and a coupon sensitivity evaluation unit for inputting the coupon attribute features into the sensitivity evaluation plug-in for evaluation, and outputting the coupon sensitivity.

[0029] Specifically, the attribute features of a virtual coupon are read through database query or API interface, including the coupon type, the product value, the remaining valid time length, and the cancellation complexity. The higher the cancellation complexity, the higher the security, so the relative storage encryption strength can be lower. The extracted attribute features are converted into a unified format, for example, the coupon type is converted into a numerical code (such as coupon = 1, gift card = 2, etc.), the product value and the remaining valid time length are kept as numerical types, and the cancellation complexity is converted into a numerical value through a preset rule (such as low complexity = 1, high complexity = 3). For example, the attribute features of a coupon read from the database are: the coupon type is "coupon", the product value is 500 yuan, the remaining valid time length is 30 days, and the cancellation complexity is "high". The coupon type "coupon" is converted into the numerical value 1, and the cancellation complexity "high" is converted into the numerical value 3. The final feature vector is: [1, 500, 30, 3].

[0030] A large number of sample coupon attribute features and corresponding sensitivity labels are collected to form a sample coupon attribute feature set and a sample coupon sensitivity set. The sample data is trained using a feedforward neural network (FNN). The feedforward neural network is a simple neural network structure, which includes an input layer, a hidden layer, and an output layer. The network weights are adjusted through the backpropagation algorithm to make the network output as close as possible to the true sensitivity label. The training process continues until the loss function (such as mean square error) of the network converges to a small value, indicating that the model has learned the mapping relationship between the coupon attribute features and the sensitivity. The trained neural network model is saved as a sensitivity evaluation plug-in.

[0031] The feature vector obtained from the card attribute feature reading unit is input into the sensitivity evaluation plug-in (i.e., a trained neural network model). The neural network model calculates the output card sensitivity level (such as low, medium, or high) based on the input feature vector through forward propagation. This implementation method automatically learns the complex relationship between card attribute features and sensitivity through the neural network model, eliminating the need for manual rule setting and improving the accuracy and efficiency of sensitivity evaluation.

[0032] The usage record acquisition module 20 is used to acquire the current usage record of the virtual card and the historical usage record of similar virtual cards.

[0033] Specifically, SQL or NoSQL database query interfaces are used to obtain current usage records of virtual vouchers and historical usage records of similar vouchers. Usage records include user ID, time of use, location of use, and amount of use. Data warehouse technologies (such as Hadoop or Spark) are used to aggregate and analyze massive amounts of historical data to generate user behavior profiles and voucher usage statistics. Real-time data processing frameworks such as Kafka or Flink are used to obtain current usage records in real time and compare them with historical data.

[0034] In one possible implementation, the usage record acquisition module 20 includes: a high-frequency information retrieval unit, configured to perform high-frequency information retrieval in a historical card and voucher operation database using the card and voucher attribute characteristics as a comparison constraint to determine multiple virtual cards and vouchers of the same type; and a historical usage record collection unit, configured to collect historical usage records of the multiple virtual cards and vouchers of the same type.

[0035] Specifically, using card attribute characteristics (such as card type, product value, remaining validity period, and redemption complexity) as comparison constraints, the historical card operation database is searched for similar virtual cards with similar attributes to the current card. Inverted index or hash index technology is used to optimize the index of card attributes in the historical card operation database to speed up retrieval. Algorithms such as cosine similarity or jaccard similarity are used to calculate the similarity between the current card attributes and historical card attributes, screening out similar cards with high similarity.

[0036] The historical usage records of similar virtual cards and coupons are collected and screened from the historical card and coupon operation database, including information such as user ID, usage time, usage location, and usage amount. The collected historical usage records are aggregated and analyzed to generate user behavior profiles and card and coupon usage statistics. For example, the average usage frequency and average usage amount of similar cards and coupons are calculated. The collected historical usage records are cached in memory or a distributed cache system (such as Redis) for fast access and analysis. In this implementation, the high-frequency information retrieval unit can quickly and accurately retrieve similar cards and coupons with similar attributes to the current card and coupon through index optimization and similarity calculation, thereby improving retrieval efficiency and matching accuracy. For example, inverted index technology can reduce retrieval time from linear complexity to logarithmic complexity, significantly improving retrieval performance.

[0037] In one possible implementation, the high-frequency information retrieval unit includes: a first feature setting unit, configured to randomly select any one feature from the card voucher type, product value, remaining validity period, and cancellation complexity without replacement and set it as a first feature; a similarity comparison unit, configured to perform similarity comparison in a historical card voucher operation database using the first feature as a comparison constraint to obtain a plurality of first virtual cards of the same type that meet a preset similarity threshold; a traversal unit, configured to continue to randomly select features without replacement for similarity comparison and screening based on the plurality of first virtual cards of the same type until the feature traversal is completed to obtain a plurality of virtual cards of the same type; and a similarity comprehensive calculation unit, configured to perform similarity comprehensive calculation on each of the plurality of virtual cards of the same type, determine a plurality of total similarities, and identify the plurality of virtual cards of the same type.

[0038] Specifically, a feature is randomly selected without replacement from among the following: coupon type, product value, remaining validity period, and redemption complexity as the first feature, and the selected first feature is recorded. Assuming the current coupon attributes are: coupon type: coupon, product value: 500 yuan, remaining validity period: 30 days, and redemption complexity: high, the first feature setting unit randomly selects "product value" as the first feature.

[0039] Using the first feature as a comparison constraint, perform a similarity comparison in the historical card and voucher operation database. Calculate the similarity between the first feature of the current card and the corresponding feature of the historical card. For example, use Euclidean distance or Manhattan distance to calculate the similarity of numerical features (such as product value), and use a string matching algorithm to calculate the similarity of categorical features (such as card and voucher type). Obtain multiple virtual cards of the first type that meet a preset similarity threshold.

[0040] For example, using "product value" as the comparison constraint and assuming a preset similarity threshold of ±10%, calculate the similarity between the current coupon (product value = 500 yuan) and the product value of historical coupons. For example, the product values ​​of historical coupons are 450 yuan, 500 yuan, and 550 yuan, all of which meet the similarity threshold (450-550 yuan). Filter out historical coupons with product values ​​between 450-550 yuan. Assume that the selected coupon IDs are 12345, 67890, and 11223.

[0041] Based on the first set of similar virtual vouchers that have been screened, the next feature is randomly selected without replacement for similarity comparison and screening. This similarity comparison process is repeated, gradually narrowing the screening range until all features have been traversed, resulting in multiple similar virtual vouchers with similarities to multiple features exceeding the similarity threshold.

[0042] For example, suppose "Remaining validity period" is randomly selected as the second feature, and the similarity threshold is preset to ±5 days. The current coupon has a remaining validity period of 30 days, and the filtered coupon IDs are 12345, 67890, and 11223. When calculating similarity, assume that coupon IDs 12345 and 67890 have a remaining validity period of 30 days and 35 days, respectively, meeting the similarity threshold (25-35 days), while coupon ID 11223 has a remaining validity period of 40 days, which does not meet the similarity threshold. Coupon IDs 12345 and 67890 are retained. Suppose "Card type" is randomly selected as the third feature. The current coupon type is "Coupon," and the filtered coupon IDs are 12345 and 67890. Assume that coupon IDs 12345 and 67890 are both of the "Coupon" type, meeting the similarity threshold. Coupon IDs 12345 and 67890 are retained. Suppose "Redeem Complexity" is randomly selected as the fourth feature. The current card / voucher's Redeem Complexity is "High," and the selected cards / vouchers have IDs 12345 and 67890. Assume that the Redeem Complexity for card / voucher 12345 is "High," while the Redeem Complexity for card / voucher 67890 is "Medium." These cards do not meet the similarity threshold. Ultimately, card / voucher ID 12345 is retained.

[0043] For the final selected similar virtual coupons, their similarity with the current coupon across multiple features is calculated and weighted summed. Weights are assigned to different features based on business needs. For example, the product value is weighted 0.4, the remaining validity period is weighted 0.3, the coupon type is weighted 0.2, and the redemption complexity is weighted 0.1. The calculated comprehensive similarity is used as an identifier and stored in the database or returned to subsequent modules.

[0044] For example, assume that the final screened coupon ID is 12345, and its similarity with the current coupon in each feature is as follows: product value similarity: 0.95, remaining valid duration similarity: 1.0, coupon type similarity: 1.0, and redemption complexity similarity: 1.0. The weight distribution is as follows: product value weight is 0.4, remaining valid duration weight is 0.3, coupon type weight is 0.2, and redemption complexity weight is 0.1. The comprehensive similarity = (0.95 x 0.4) + (1.0 x 0.3) + (1.0 x 0.2) + (1.0 x 0.1) = 0.98. This implementation gradually screens multiple features to ensure that the final screened virtual coupon of the same type is highly similar to the current coupon in multiple dimensions, thereby improving the accuracy of the match.

[0045] The user redemption probability prediction module 30 is configured to predict the user redemption probability in a preset time zone according to the user's coupon usage behavior characteristics, the current usage record, and the historical usage record, and output the predicted user redemption probability.

[0046] Specifically, features are extracted from user behavior data, such as the user's historical redemption frequency, redemption time interval, and redemption amount distribution. A data-driven prediction method is used to predict the user redemption probability. The model outputs the user's redemption probability in a future preset time zone (such as a week) according to the user's coupon usage behavior characteristics, the current usage record, and the historical usage record of the same type of virtual coupon.

[0047] In one possible implementation, the user redemption probability prediction module 30 includes: a redemption proportion acquisition unit configured to acquire the redemption proportions of a plurality of virtual coupons of the same type in a preset historical time zone according to the user's coupon usage behavior characteristics, and set the plurality of self-sample redemption probabilities, thereby obtaining a plurality of self-sample redemption probabilities; a first predicted redemption probability output unit configured to configure a credible weight proportion according to the plurality of total similarities, weight and fuse the plurality of self-sample redemption probabilities, and output a first predicted redemption probability; a second predicted redemption probability output unit configured to predict the user redemption probability according to the current usage record, and output a second predicted redemption probability; a third predicted redemption probability output unit configured to predict the user redemption probability according to the historical usage record, and output a third predicted redemption probability; and an evaluation and determination unit configured to evaluate and determine the predicted user redemption probability based on the first predicted redemption probability, the second predicted redemption probability, and the third predicted redemption probability.

[0048] Specifically, according to the user's coupon use behavior characteristics, the redemption records of multiple similar virtual coupons in the preset historical time zone are extracted from the historical coupon operation database. The redemption proportion of each similar virtual coupon, i.e. the ratio of the redemption times to the issuance times, is calculated and set as the self-sample redemption probability of the similar coupon. The calculated multiple self-sample redemption probabilities are stored as a list or array.

[0049] According to the total similarity of multiple similar virtual coupons (provided by the high-frequency information retrieval unit), a trusted weight proportion is configured for each similar coupon. The higher the total similarity, the greater the weight proportion. The multiple self-sample redemption probabilities and the corresponding trusted weight proportions are weighted and summed to output the first predicted redemption probability. For example, assuming that the total similarity of coupon ID 12345 is 0.98 and the total similarity of coupon ID 67890 is 0.92. According to the total similarity, the weight proportion of coupon ID 12345 is 0.98 / (0.98+0.92)=0.516, and the weight proportion of coupon ID 67890 is 0.92 / (0.98+0.92)=0.484.

[0050] The real-time behavior characteristics of the user are extracted from the current use record, such as the current use frequency, the use amount, the use time, etc. The data-driven prediction method is used to predict the redemption probability of the current use record, and the second predicted redemption probability is output.

[0051] The long-term behavior characteristics of the user are extracted from the historical use record, such as the average use frequency, the average use amount, the historical redemption rate, etc. The data-driven prediction method is used to predict the redemption probability of the historical use record, and the third predicted redemption probability is output.

[0052] The first predicted redemption probability, the second predicted redemption probability and the third predicted redemption probability are comprehensively evaluated to determine the final predicted user redemption probability. Each predicted redemption probability is assigned a weight, and the weight distribution is adjusted according to business needs. For example, the first predicted redemption probability weight is 0.5, the second predicted redemption probability weight is 0.3, and the third predicted redemption probability weight is 0.2. This implementation method predicts the user redemption probability from multiple dimensions by combining the redemption proportion of similar coupons, the current use record and the historical use record, thereby improving the accuracy and reliability of the user redemption probability prediction.

[0053] In a possible implementation, the second prediction cancellation probability output unit comprises: a second initial prediction cancellation probability setting unit configured to, according to the current use record, statistically determine a cancellation proportion of the virtual card within a preset historical time zone, and set the cancellation proportion as a second initial prediction cancellation probability; a final cancellation proportion determination unit configured to, based on a historical card operation database, collect a plurality of final cancellation proportions of a plurality of similar virtual cards, and determine a final cancellation proportion by weighting according to the credible weight proportion; a unit standard cancellation progress setting unit configured to obtain a card validity duration of the virtual card, set a ratio of the final cancellation proportion to the card validity duration as a unit standard cancellation progress; a real-time standard cancellation progress setting unit configured to, according to the current use record, obtain a cancelled proportion of the virtual card, subtract the cancelled proportion from the final cancellation proportion to obtain a predicted uncancellation proportion, and set a ratio of the predicted uncancellation proportion to a remaining validity duration as a real-time standard cancellation progress; and a correction unit configured to set a ratio of the real-time standard cancellation progress to the unit standard cancellation progress as a correction coefficient, correct the second initial prediction cancellation probability, and output the second prediction cancellation probability.

[0054] Specifically, according to the current use record, a cancellation record of the virtual card within a preset historical time zone is statistically determined. A ratio of a cancellation number to an issuance number is calculated and set as a second initial prediction cancellation probability. For example, assuming that the preset historical time zone is the past 3 months, the cancellation record of the current virtual card is extracted, the total issuance number is 100 times, and the cancellation number is 70 times, the second initial prediction cancellation probability is 0.7 (70%).

[0055] The final cancellation proportions of a plurality of similar virtual cards are collected from the historical card operation database. The final cancellation proportions are weighted and averaged according to the credible weight proportion (determined by the total similarity) to determine the final cancellation proportion. For example, assuming that the final cancellation proportions of the collected similar cards are 0.65, 0.70 and 0.68 respectively, and the credible weight proportions are 0.4, 0.3 and 0.3 respectively, the final cancellation proportion is (0.65*0.4) + (0.70*0.3) + (0.68*0.3) = 0.67.

[0056] The validity duration of the virtual card is obtained, and a ratio of the final cancellation proportion to the card validity duration is set as a unit standard cancellation progress. For example, assuming that the validity duration of the virtual card is 60 days, the unit standard cancellation progress = 0.67 / 60 ≈ 0.0112 (the cancellation progress per day).

[0057] Based on the current usage history, obtain the redeemed percentage of the virtual voucher. Subtract the redeemed percentage from the final redeemed percentage to obtain the predicted unredeemed percentage. Set the ratio of the predicted unredeemed percentage to the remaining valid duration as the real-time standard redemption progress. For example, assuming the current redeemed percentage is 0.4 (40%), then the predicted unredeemed percentage = 0.67 - 0.4 = 0.27. Assuming the current remaining valid duration is 30 days, then the real-time standard redemption progress = 0.27 / 30 = 0.009.

[0058] The ratio of the real-time standard write-off progress to the unit standard write-off progress is set as the correction coefficient. The second initial predicted write-off probability is corrected using the correction coefficient, and the second predicted write-off probability is output. Continuing with the above example, the correction coefficient = 0.009 / 0.0112≈0.8036, and the second predicted write-off probability = 0.7×0.8036≈0.5625. This implementation method dynamically adjusts the second initial predicted write-off probability through the correction unit, taking into account the differences between current usage records and historical data, thereby improving the accuracy of the prediction. For example, if the current write-off progress is lower than expected, the correction coefficient will reduce the second initial predicted write-off probability to make it closer to the actual situation.

[0059] In one possible implementation, the third predicted write-off probability output unit includes: a sample write-off probability acquisition unit, which is used to count the write-off ratio of virtual cards and coupons in a preset historical time zone based on the historical usage records, and obtain multiple sample write-off probabilities; a weighted fusion unit, which is used to weightedly fuse the multiple sample write-off probabilities according to the trust weight ratio, and output the third predicted write-off probability.

[0060] Specifically, extract the virtual voucher redemption records within a preset historical time zone from historical usage records, including the number of redemptions and issuances. Calculate the redemption percentage for each virtual voucher—that is, the ratio of redemptions to issuances—as a sample redemption probability. Store the calculated sample redemption probabilities in a list or array.

[0061] Each sample's write-off probability is weighted based on its credibility weight (the total similarity provided by the high-frequency information retrieval unit). The weighted sum of multiple sample write-off probabilities and their corresponding credibility weights is then calculated to output a third predicted write-off probability. This implementation method uses the statistical properties of historical data to predict write-off probabilities by counting write-off percentages from historical usage records, ensuring a highly stable and reliable prediction.

[0062] The encryption storage module 40 is configured to determine the card encryption strength based on the card sensitivity and the predicted user redemption probability, and encrypt and store the virtual card in the preset time zone.

[0063] Specifically, the appropriate encryption algorithm is dynamically selected based on the sensitivity of the card or voucher and the predicted probability of user redemption. For example, high-sensitivity cards or vouchers are encrypted using AES-256, medium-sensitivity cards or vouchers are encrypted using AES-128, and low-sensitivity cards or vouchers are encrypted using AES-128 but with a lower encryption frequency. For cards or vouchers with a higher predicted probability of user redemption, the encryption frequency or strength is appropriately increased to ensure a higher level of protection for high-value or frequently used card or voucher data. A key management system (KMS) is used to generate, store, and manage encryption keys. Keys can be generated dynamically or adjusted based on user behavior and card or voucher attributes. Based on encryption strength and performance requirements, appropriate storage media, such as SSDs or encrypted storage chips (such as TPMs), are selected.

[0064] In one possible implementation, the encryption storage module 40 includes: an initial card / voucher encryption strength matching unit, configured to determine the initial card / voucher encryption strength based on the card / voucher sensitivity matching; a strength adjustment coefficient setting unit, configured to set the ratio of the predicted user redemption probability to the unit standard redemption progress of the virtual card / voucher as the strength adjustment coefficient; and an encryption strength optimization unit, configured to optimize the initial card / voucher encryption strength based on the strength adjustment coefficient and output the card / voucher encryption strength.

[0065] Specifically, cards are categorized into three levels based on their sensitivity: high, medium, and low. The initial encryption strength is matched to the card's sensitivity level. For example, high-sensitivity cards use AES-256 encryption every hour, medium-sensitivity cards use AES-128 every four hours, and low-sensitivity cards use AES-128 every 12 hours. For example, if a card's sensitivity is "medium," its initial encryption strength is AES-128, with an encryption frequency of every four hours.

[0066] The ratio of the predicted user redemption probability to the unit standard redemption progress of the virtual card is set as the strength adjustment coefficient. For example, if the predicted user redemption probability of a card is 0.7 and the unit standard redemption progress is 0.0112 (daily redemption progress), the strength adjustment coefficient = 0.7 / 0.0112 = 62.5.

[0067] The initial pass / card encryption strength is optimized based on the strength adjustment factor. A higher strength adjustment factor indicates that the predicted user redemption probability is significantly higher than the unit's standard redemption rate, necessitating an increase in encryption strength or frequency. The encryption strength optimization rule could be: if the strength adjustment factor is >100, the encryption algorithm is upgraded to AES-256, and the encryption frequency is increased to once an hour; if the strength adjustment factor is between 50 and 100, the encryption frequency is increased to once every two hours; if the strength adjustment factor is <50, the initial encryption strength remains unchanged. Continuing with the above example, if the pass / card strength adjustment factor is 62.5, the optimization rule would increase the encryption frequency from once every four hours to once every two hours, while maintaining AES-128 as the encryption algorithm. This implementation dynamically adjusts encryption strength using the strength adjustment factor, allowing for flexible adjustments to encryption policies based on changes in the predicted user redemption probability, ensuring higher levels of protection for passes / cards with a high redemption probability. For example, for passes / cards with a higher redemption probability, the encryption frequency and algorithm strength are automatically increased to address higher security risks. For cards and coupons with a low probability of redemption, maintain a low encryption frequency to avoid unnecessary encryption operations and save computing resources and storage costs.

[0068] The embodiments of the present application use technical means such as evaluating the sensitivity of virtual cards and predicting the probability of user redemption based on the card attribute characteristics, current and historical usage records of the virtual cards, and then analyzing and determining the encryption strength of different cards and performing differentiated encrypted storage. This solves the technical problem of the existing encrypted storage of virtual cards and the difficulty in balancing storage security and redemption efficiency, achieving the technical effect of improving redemption efficiency while ensuring the storage security of virtual cards and coupons.

[0069] In the above, refer to Figure 1 The hierarchical encryption storage system of virtual cards and coupons according to the embodiment of the present invention is described in detail. Figure 2 A hierarchical encryption storage method for virtual cards according to an embodiment of the present invention is described.

[0070] The hierarchical encrypted storage method for virtual cards and coupons according to an embodiment of the present invention is used to solve the technical problem of balancing storage security and redemption efficiency in existing encrypted storage of virtual cards and coupons, thereby achieving the technical effect of improving redemption efficiency while ensuring the storage security of virtual cards and coupons.

[0071] The hierarchical encryption storage method of the virtual coupon comprises: evaluating coupon sensitivity according to coupon attribute characteristics of the virtual coupon, and outputting coupon sensitivity; obtaining current use records of the virtual coupon and historical use records of similar virtual coupons; predicting user cancellation probability in a preset time zone according to user coupon use behavior characteristics, the current use records and the historical use records, and outputting predicted user cancellation probability; analyzing and determining coupon encryption strength according to the coupon sensitivity and the predicted user cancellation probability, and performing encryption storage on the virtual coupon in the preset time zone.

[0072] The coupon sensitivity evaluation according to the coupon attribute characteristics of the virtual coupon can further comprise: reading the coupon attribute characteristics of the virtual coupon, wherein the coupon attribute characteristics at least include coupon type, commodity value, remaining valid time length and cancellation complexity; training a feedforward neural network to convergence using a sample coupon attribute characteristic set and a sample coupon sensitivity set, and obtaining a sensitivity evaluation plug-in; inputting the coupon attribute characteristics into the sensitivity evaluation plug-in for evaluation, and outputting the coupon sensitivity.

[0073] The obtaining of the historical use records of similar virtual coupons can further comprise: performing high-frequency information retrieval in a historical coupon operation database with the coupon attribute characteristics as a comparison constraint, and determining a plurality of similar virtual coupons; and collecting historical use records of the plurality of similar virtual coupons.

[0074] The determination of the plurality of similar virtual coupons can further comprise: randomly selecting any feature in the coupon type, commodity value, remaining valid time length and cancellation complexity as a first feature without replacement; performing similarity comparison in a historical coupon operation database with the first feature as a comparison constraint, and obtaining a plurality of first similar virtual coupons satisfying a preset similarity threshold; based on the plurality of first similar virtual coupons, continuing to randomly select features without replacement for similarity comparison and screening until the features are traversed, and obtaining a plurality of similar virtual coupons; respectively performing similarity degree comprehensive calculation on the plurality of similar virtual coupons, determining a plurality of total similarities, and identifying the plurality of similar virtual coupons.

[0075] The method can further include: obtaining a plurality of self-sample redemption probabilities of the same type of virtual coupons in a preset historical time zone according to the card coupon use behavior characteristics of the user, and setting the plurality of self-sample redemption probabilities as self-sample redemption probabilities; weighting and fusing the plurality of self-sample redemption probabilities according to the plurality of total similarity configuration credible weight proportions to output a first predicted redemption probability; predicting a user redemption probability according to the current use record to output a second predicted redemption probability; predicting a user redemption probability according to the historical use record to output a third predicted redemption probability; and determining the predicted user redemption probability based on the first predicted redemption probability, the second predicted redemption probability, and the third predicted redemption probability.

[0076] The method can further include: obtaining a plurality of self-sample redemption probabilities of the same type of virtual coupons in a preset historical time zone according to the card coupon use behavior characteristics of the user, and setting the plurality of self-sample redemption probabilities as self-sample redemption probabilities; weighting and fusing the plurality of self-sample redemption probabilities according to the plurality of total similarity configuration credible weight proportions to output a first predicted redemption probability; predicting a user redemption probability according to the current use record to output a second predicted redemption probability; predicting a user redemption probability according to the historical use record to output a third predicted redemption probability; and determining the predicted user redemption probability based on the first predicted redemption probability, the second predicted redemption probability, and the third predicted redemption probability.

[0077] The method can further include: obtaining a plurality of self-sample redemption probabilities of the same type of virtual coupons in a preset historical time zone according to the card coupon use behavior characteristics of the user, and setting the plurality of self-sample redemption probabilities as self-sample redemption probabilities; weighting and fusing the plurality of self-sample redemption probabilities according to the plurality of total similarity configuration credible weight proportions to output a first predicted redemption probability; predicting a user redemption probability according to the current use record to output a second predicted redemption probability; predicting a user redemption probability according to the historical use record to output a third predicted redemption probability; and determining the predicted user redemption probability based on the first predicted redemption probability, the second predicted redemption probability, and the third predicted redemption probability.

[0078] The method can further include: obtaining a plurality of self-sample redemption probabilities of the same type of virtual coupons in a preset historical time zone according to the card coupon use behavior characteristics of the user, and setting the plurality of self-sample redemption probabilities as self-sample redemption probabilities; weighting and fusing the plurality of self-sample redemption probabilities according to the plurality of total similarity configuration credible weight proportions to output a first predicted redemption probability; predicting a user redemption probability according to the current use record to output a second predicted redemption probability; predicting a user redemption probability according to the historical use record to output a third predicted redemption probability; and determining the predicted user redemption probability based on the first predicted redemption probability, the second predicted redemption probability, and the third predicted redemption probability.

[0079] The hierarchical encryption storage system of the virtual card provided by the embodiment of the present application can execute the hierarchical encryption storage method of the virtual card provided by any embodiment of the present application, and has the function modules and beneficial effects corresponding to the execution method.

[0080] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or the server, and the various units and modules are only divided according to the functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for the convenience of mutual differentiation, and are not used to limit the protection scope of the present application.

[0081] Based on the foregoing embodiments, the embodiment of the present application further provides a computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program is executed by a processor of an electronic device, and can realize the method according to any one of the foregoing embodiments.

[0082] The specific embodiments described above do not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps described in the present application can be executed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.

Claims

1. A hierarchical encrypted storage system for virtual cards and coupons, characterized by: The hierarchical encryption storage system includes: A card / voucher sensitivity evaluation module is used to evaluate the card / voucher sensitivity based on the card / voucher attribute characteristics of the virtual card / voucher and output the card / voucher sensitivity; A usage record acquisition module is used to acquire the current usage record of the virtual card and the historical usage record of similar virtual cards; A user redemption probability prediction module is used to predict the user redemption probability within a preset time zone based on the user's card or voucher usage behavior characteristics, the current usage record, and the historical usage record, and output the predicted user redemption probability; The encryption storage module is used to determine the card encryption strength based on the card sensitivity and the predicted user cancellation probability analysis, and encrypt and store the virtual card in the preset time zone.

2. The hierarchical encryption storage system for virtual cards and coupons according to claim 1, characterized in that: The card and coupon sensitivity evaluation module includes: A card attribute reading unit, configured to read the card attribute characteristics of a virtual card, wherein the card attribute characteristics include at least the card type, commodity value, remaining validity period, and redemption complexity; A sensitivity evaluation plug-in acquisition unit is used to train a feedforward neural network using a sample card attribute feature set and a sample card sensitivity set until convergence, and to acquire a sensitivity evaluation plug-in; The card / voucher sensitivity evaluation unit is configured to input the card / voucher attribute characteristics into the sensitivity evaluation plug-in for evaluation and output the card / voucher sensitivity.

3. The hierarchical encryption storage system for virtual cards and coupons according to claim 2, characterized in that: The usage record acquisition module includes: A high-frequency information retrieval unit, configured to perform a high-frequency information search in a historical card and voucher operation database using the card and voucher attribute characteristics as a comparison constraint to determine a plurality of similar virtual cards and vouchers; The historical usage record collecting unit is used to collect historical usage records of the multiple virtual cards of the same type.

4. The hierarchical encryption storage system for virtual cards and coupons according to claim 3, characterized in that: The high-frequency information retrieval unit includes: A first feature setting unit is configured to randomly select any one of the card or voucher type, product value, remaining validity period, and redemption complexity without replacement and set it as the first feature; a similarity comparison unit, configured to perform a similarity comparison in a historical card and voucher operation database using the first feature as a comparison constraint, and obtain a plurality of virtual cards and vouchers of the first type that meet a preset similarity threshold; A traversal unit is configured to continue randomly selecting features without replacement to perform similarity comparison and screening based on the plurality of first virtual cards of the same type, until the feature traversal is completed to obtain a plurality of virtual cards of the same type; The similarity comprehensive calculation unit is used to perform comprehensive similarity calculations on the multiple virtual cards of the same type, determine multiple total similarities, and identify the multiple virtual cards of the same type.

5. The hierarchical encryption storage system for virtual cards and coupons according to claim 4, characterized in that: The user write-off probability prediction module includes: A write-off ratio obtaining unit is used to obtain the write-off ratio of multiple virtual cards and coupons of the same type in a preset historical time zone based on the user's card and coupon usage behavior characteristics, set the write-off ratio as the self-sample write-off ratio, and obtain multiple self-sample write-off ratios; A first predicted cancellation probability output unit is configured to configure a trust weight ratio according to the multiple total similarities, perform weighted fusion on the multiple self-sample cancellation probabilities, and output a first predicted cancellation probability; a second predicted write-off probability output unit, configured to predict the user write-off probability based on the current usage record and output a second predicted write-off probability; a third predicted write-off probability output unit, configured to predict the user write-off probability based on the historical usage record and output a third predicted write-off probability; An evaluation and determination unit is configured to evaluate and determine a predicted user write-off probability based on the first predicted write-off probability, the second predicted write-off probability, and the third predicted write-off probability.

6. The hierarchical encryption storage system for virtual cards and coupons according to claim 5, characterized in that: The second predicted write-off probability output unit includes: a second initial predicted write-off probability setting unit, configured to calculate a write-off ratio of virtual coupons in a preset historical time zone based on the current usage record, and set the ratio as a second initial predicted write-off probability; A final write-off ratio determination unit is configured to collect, based on a historical card and coupon operation database, multiple average final write-off ratios of multiple similar virtual cards and coupons, and determine the final write-off ratio by weighting them according to the trust weight ratio; a unit standard redemption progress setting unit, configured to obtain the card coupon validity period of the virtual card coupon, and set the unit standard redemption progress as the ratio of the final redemption ratio to the card coupon validity period; a real-time standard redemption progress setting unit, configured to obtain a redemption ratio of the virtual card or voucher based on the current usage record, subtract the redemption ratio from the final redemption ratio to obtain a predicted unredeemed ratio, and set the ratio of the predicted unredeemed ratio to the remaining validity period as the real-time standard redemption progress; The correction unit is configured to set a ratio of the real-time standard write-off progress to the unit standard write-off progress as a correction coefficient, correct the second initial predicted write-off probability, and output the second predicted write-off probability.

7. The hierarchical encryption storage system for virtual cards and coupons according to claim 5, characterized in that: The third predicted write-off probability output unit includes: a sample redemption probability acquisition unit, configured to calculate the redemption ratio of virtual cards and coupons in a preset historical time zone based on the historical usage records, and acquire multiple sample redemption probabilities; A weighted fusion unit is used to perform weighted fusion on the multiple sample write-off probabilities according to the credibility weight ratio and output the third predicted write-off probability.

8. The hierarchical encryption storage system for virtual cards and coupons according to claim 6, characterized in that: The encryption storage module includes: an initial card / voucher encryption strength matching unit, configured to determine the initial card / voucher encryption strength according to the card / voucher sensitivity matching; a strength adjustment coefficient setting unit, configured to set a ratio of the predicted user redemption probability to the unit standard redemption progress of the virtual card or voucher as a strength adjustment coefficient; The encryption strength optimization unit is used to optimize the initial card and coupon encryption strength according to the strength adjustment coefficient and output the card and coupon encryption strength.

9. A hierarchical encryption storage method for virtual cards and coupons, characterized in that: The method is implemented by the hierarchical encryption storage system for virtual cards according to any one of claims 1 to 8, and the method includes: Evaluate the card sensitivity of the virtual card based on its attribute characteristics and output the card sensitivity; Obtaining the current usage record of the virtual card and the historical usage record of similar virtual cards; Predicting the user's redemption probability within a preset time zone based on the user's card or voucher usage behavior characteristics, the current usage record, and historical usage records, and outputting the predicted user redemption probability; The card voucher encryption strength is determined based on the card voucher sensitivity and the predicted user redemption probability, and the virtual card voucher within the preset time zone is encrypted and stored.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the hierarchical encryption storage method for virtual cards and coupons as claimed in claim 9 is implemented.

Images