Mobile terminal secure payment management method and system

By obtaining and processing transaction payment data of different risk orders, generating prototype characterization feature vectors, intelligently identifying the risk types of transaction data to be paid and encrypted, the problems of multimodal data processing efficiency and model robustness in the existing technology are solved. , improves the efficiency and accuracy of mobile terminal security payment management.

CN118333626BActive Publication Date: 2024-11-22SHENZHEN SHICHUANG INTELLIGENT TECH CO LTD
View PDF 10 Cites 0 Cited by

Patent Information

Application Number
CN202410489303.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-04-23
Publication Date
2024-11-22
Estimated Expiration
2044-04-23

AI Technical Summary

Technical Problem

The existing mobile terminal security payment management method consumes a lot of computing resources when dealing with multimodal data conversion and semantic alignment, and the robustness and generalization capabilities of the model are difficult to guarantee.

Method used

By obtaining the transaction payment data set marked as low, medium and high risk orders, abstract representation of semantic features, generating prototype characterization feature vectors, and determining the risk type tag of transaction data to be paid based on these vectors for data encryption, using gate control Switch prototype feature screening network optimizes feature characterization to improve the robustness and generalization ability of the model.

Benefits of technology

It realizes intelligent identification of the risk types of transaction data to be paid, makes full use of the data structure consistency of the sample set, improves the robustness and generalization ability of model training, and reduces the consumption of computing resources.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118333626B_ABST
    Figure CN118333626B_ABST
Patent Text Reader

Abstract

The present disclosure relates to a method and system for managing secure payment of a mobile terminal. The method includes: obtaining a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, a set of third transaction payment data marked as high-risk orders, and transaction data to be paid, and characterizing the transaction risk type of the transaction data to be paid by quantifying the similarity between the semantic information in the transaction data to be paid and the semantic risk information in the transaction payment data marked by the data in the feature common space, thereby intelligently identifying the risk type of the transaction data to be paid. In this way, the relatively good consistency of the data structure of the sample set can be fully utilized to improve the robustness and generalization ability of model training.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the field of secure payment technology, and in particular, to a mobile terminal secure payment management method and system. Background Art

[0002] Near Field Communication (NFC) provides greater convenience and speed in the field of mobile payment, but it also faces challenges in data security, such as the risk of data being stolen or intercepted. In order to improve payment security, biometric technology has been introduced into mobile payment management. Technologies such as fingerprint recognition, facial recognition and iris scanning have gradually become ways to verify payment identities and are widely used because of their uniqueness and difficulty in forging.

[0003] The rise of encryption technology and blockchain has also brought more innovation to mobile payments. Encryption technology ensures the secure transmission and storage of data, while blockchain technology provides a decentralized secure payment solution, enhancing the transparency and security of the payment system. However, the existing technologies for transaction orders with different risks often use a single encryption method, and often lack consideration of user privacy and environmental credibility during the payment process.

[0004] In response to the above problems, patent CN 117557270A provides a mobile terminal secure payment management method and system, which quantifies and evaluates transaction risks of transaction data, and adopts different encryption methods to encapsulate data based on transaction orders of different risk levels to generate a secure transaction payment protocol. At the same time, it uses blockchain technology and privacy protection technology to enhance the security and privacy of the mobile payment system.

[0005] However, there are some problems in the actual application of the above-mentioned mobile terminal security payment management method. For example, in the process of using the payment behavior biometric map to train the model of the transaction data to be paid to generate a transaction risk assessment prediction model, it is necessary to deal with multimodal data conversion and semantic alignment problems, which not only consumes a lot of computing resources, but also makes it difficult to ensure the robustness and generalization ability of the model.

[0006] Therefore, an optimized mobile terminal secure payment management method and system are desired. Summary of the invention

[0007] This summary is provided to introduce concepts in a brief form that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0008] In a first aspect, the present disclosure provides a method for managing secure payment on a mobile terminal, the method comprising:

[0009] Acquire a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, and a set of third transaction payment data marked as high-risk orders;

[0010] Performing transaction payment data semantic feature abstract characterization on the set of first transaction payment data marked as low-risk orders, the set of second transaction payment data marked as medium-risk orders, and the set of third transaction payment data marked as high-risk orders to obtain a low-risk order transaction payment data prototype characterization feature vector, a medium-risk order transaction payment data prototype characterization feature vector, and a high-risk order transaction payment data prototype characterization feature vector;

[0011] Get the transaction data to be paid;

[0012] Determine the risk type label of the transaction data to be paid based on the similarity between the transaction payment semantic feature information of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data, and the prototype characterization feature vector of the high-risk order transaction payment data;

[0013] Data encryption is performed based on the risk type label of the transaction data to be paid.

[0014] Optionally, the set of first transaction payment data marked as low-risk orders, the set of second transaction payment data marked as medium-risk orders, and the set of third transaction payment data marked as high-risk orders are subjected to transaction payment data semantic feature abstract representation to obtain low-risk order transaction payment data prototype representation feature vectors, medium-risk order transaction payment data prototype representation feature vectors, and high-risk order transaction payment data prototype representation feature vectors, including: semantically encoding each first transaction payment data marked as low-risk order in the set of first transaction payment data marked as low-risk orders to obtain a set of low-risk order transaction payment data semantic encoding feature vectors; passing the set of low-risk order transaction payment data semantic encoding feature vectors through a gated switch prototype feature screening network to obtain the low-risk order transaction payment data prototype representation feature vector; and encoding the set of second transaction payment data marked as medium-risk orders and the set of third transaction payment data marked as high-risk orders in the same manner to obtain the medium-risk order transaction payment data prototype representation feature vectors and the high-risk order transaction payment data prototype representation feature vectors.

[0015] Optionally, the set of semantically encoded feature vectors of the low-risk order transaction payment data is passed through a gated switch prototype feature screening network to obtain the prototype representation feature vector of the low-risk order transaction payment data, including: processing the set of semantically encoded feature vectors of the low-risk order transaction payment data with the following gated switch prototype feature screening formula to obtain the prototype representation feature vector of the low-risk order transaction payment data; wherein the gated switch prototype feature screening formula is:

[0016]

[0017]

[0018]

[0019] Among them, v i is the i-th low-risk order transaction payment data semantic encoding feature vector in the set of low-risk order transaction payment data semantic encoding feature vectors, v j is the jth low-risk order transaction payment data semantic encoding feature vector in the set of low-risk order transaction payment data semantic encoding feature vectors, ||·||1 represents the 1-norm of the feature vector, M is the length of the set of low-risk order transaction payment data semantic encoding feature vectors -1, V k The representation of the set of semantically encoded feature vectors of the low-risk order transaction payment data, D i represents the feature difference coefficient, mask represents masking processing, represents the feature difference coefficient after masking, exp(·) represents the natural exponential function operation, N s represents the total number of feature difference coefficients after masking, P k A feature vector is provided for representing a prototype of payment data for the low-risk order transaction.

[0020] Optionally, the set of second transaction payment data marked as medium-risk orders and the set of third transaction payment data marked as high-risk orders are encoded in the same manner to obtain the medium-risk order transaction payment data prototype characterization feature vector and the high-risk order transaction payment data prototype characterization feature vector, including: semantically encoding each second transaction payment data marked as medium-risk order in the set of second transaction payment data marked as medium-risk orders to obtain a set of medium-risk order transaction payment data semantic encoding feature vectors; passing the set of medium-risk order transaction payment data semantic encoding feature vectors through the gated switch prototype feature screening network to obtain the medium-risk order transaction payment data prototype characterization feature vector; semantically encoding each third transaction payment data marked as high-risk order in the set of third transaction payment data marked as high-risk orders to obtain a set of high-risk order transaction payment data semantic encoding feature vectors; passing the set of high-risk order transaction payment data semantic encoding feature vectors through the gated switch prototype feature screening network to obtain the high-risk order transaction payment data prototype characterization feature vector.

[0021] Optionally, the risk type label of the transaction data to be paid is determined based on the degree of similarity between the transaction payment semantic feature information of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data and the prototype characterization feature vector of the high-risk order transaction payment data, including: semantically encoding the transaction data to be paid to obtain the semantic encoding feature vector of the transaction data to be paid; calculating the hash similarity between the semantic encoding feature vector of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data and the prototype characterization feature vector of the high-risk order transaction payment data to obtain a first hash similarity, a second hash similarity and a third hash similarity; based on the first hash similarity, the second hash similarity and the third hash similarity, determining the risk type label of the transaction data to be paid.

[0022] Optionally, calculating the hash similarity between the semantic encoding feature vector of the transaction data to be paid and the prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the medium-risk order transaction payment data and the prototype representation feature vector of the high-risk order transaction payment data to obtain a first hash similarity, a second hash similarity and a third hash similarity, including: using a hash mapping function to map the semantic encoding feature vector of the transaction data to be paid, the prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the medium-risk order transaction payment data and the prototype representation feature vector of the high-risk order transaction payment data to the semantic encoding hash code of the transaction data to be paid, the prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the medium-risk order transaction payment data and the prototype representation feature vector of the high-risk order transaction payment data respectively. The method comprises the following steps: calculating the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype representation hash code of the low-risk order transaction payment data; calculating the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype representation hash code of the medium-risk order transaction payment data; calculating the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype representation hash code of the medium-risk order transaction payment data; calculating the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype representation hash code of the high-risk order transaction payment data; and calculating the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype representation hash code of the high-risk order transaction payment data.

[0023] Optionally, data encryption is performed based on the risk type label of the transaction data to be paid, including: determining an encryption method for the transaction data to be paid based on the risk type label of the transaction data to be paid; and encrypting the transaction data to be paid based on the encryption method.

[0024] Optionally, the method further includes a training step: training the gate-controlled switch prototype feature screening network.

[0025] Optionally, the training step includes: obtaining training data, the training data including a set of training first transaction payment data marked as low-risk orders, a set of training second transaction payment data marked as medium-risk orders, a set of training third transaction payment data marked as high-risk orders, and a true value of the first hash similarity, a true value of the second hash similarity, and a true value of the third hash similarity; semantically encoding each training first transaction payment data marked as low-risk orders in the set of training first transaction payment data marked as low-risk orders to obtain a set of semantic encoding feature vectors of training low-risk order transaction payment data; passing the set of semantic encoding feature vectors of training low-risk order transaction payment data through the gated switch prototype feature screening network to obtain a prototype representation feature vector of training low-risk order transaction payment data; in the same way, encoding the set of training second transaction payment data marked as medium-risk orders and the set of training third transaction payment data marked as high-risk orders to obtain a prototype representation feature vector of training medium-risk order transaction payment data and a prototype representation feature vector of training high-risk order transaction payment data; obtaining training transaction to-be-paid data According to the present invention, the present invention relates to a method for detecting a gate-controlled switch prototype feature screening network. The method comprises the following steps: performing semantic encoding on the training transaction data to be paid to obtain a semantic encoding feature vector of the training transaction data to be paid; performing feature distribution optimization on the training low-risk order transaction payment data prototype representation feature vector to obtain an optimized low-risk order transaction payment data prototype representation feature vector; calculating the hash similarity between the semantic encoding feature vector of the training transaction data to be paid and the optimized low-risk order transaction payment data prototype representation feature vector, the training medium-risk order transaction payment data prototype representation feature vector and the training high-risk order transaction payment data prototype representation feature vector to obtain a training first hash similarity, a training second hash similarity and a training third hash similarity; respectively calculating the cross entropy values ​​between the training first hash similarity and the true value of the first hash similarity, between the training second hash similarity and the true value of the second hash similarity, and between the training third hash similarity and the true value of the third hash similarity to obtain a first cross entropy value, a second cross entropy value and a third cross entropy value; and taking the weighted sum of the first cross entropy value, the second cross entropy value and the third cross entropy value as the loss function value to train the gate-controlled switch prototype feature screening network.

[0026] In a second aspect, the present disclosure provides a mobile terminal secure payment management system, the system comprising:

[0027] A payment data acquisition module, used to acquire a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, and a set of third transaction payment data marked as high-risk orders;

[0028] A transaction payment data semantic feature abstract characterization module is used to perform transaction payment data semantic feature abstract characterization on the set of first transaction payment data marked as low-risk orders, the set of second transaction payment data marked as medium-risk orders, and the set of third transaction payment data marked as high-risk orders to obtain a low-risk order transaction payment data prototype characterization feature vector, a medium-risk order transaction payment data prototype characterization feature vector, and a high-risk order transaction payment data prototype characterization feature vector;

[0029] A transaction data acquisition module for obtaining transaction data to be paid, used to obtain transaction data to be paid;

[0030] A similarity determination module, configured to determine a risk type label of the transaction data to be paid based on the similarity between the transaction payment semantic feature information of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data, and the prototype characterization feature vector of the high-risk order transaction payment data;

[0031] The data encryption judgment module is used to encrypt the data based on the risk type label of the transaction data to be paid.

[0032] By adopting the above technical solution, by obtaining a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, a set of third transaction payment data marked as high-risk orders, and transaction data to be paid, the transaction risk type of the transaction data to be paid is characterized by quantifying the similarity between the semantic information in the transaction data to be paid and the semantic risk information in the transaction payment data marked by the data in the feature common space, thereby intelligently identifying the risk type of the transaction data to be paid. In this way, the relatively good consistency of the data structure of the sample set can be fully utilized to improve the robustness and generalization ability of model training.

[0033] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale. In the drawings:

[0035] Figure 1 The present invention is a flowchart of a method for managing secure payment of a mobile terminal according to an exemplary embodiment.

[0036] Figure 2 The present invention is a block diagram of a mobile terminal secure payment management system according to an exemplary embodiment.

[0037] Figure 3 It is a block diagram of an electronic device according to an exemplary embodiment.

[0038] Figure 4 It is an application scenario diagram of a mobile terminal secure payment management method according to an exemplary embodiment. DETAILED DESCRIPTION

[0039] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0040] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0041] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0042] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0043] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0044] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0045] Patent CN 117557270A provides a mobile terminal security payment management method and system, which quantifies and evaluates transaction risks of transaction data, and adopts different encryption methods to encapsulate data based on transaction orders of different risk levels to generate a secure transaction payment protocol, and uses blockchain technology and privacy protection technology to enhance the security and privacy of the mobile payment system. More specifically, during the operation of the above-mentioned mobile terminal security payment management method, it generates a payment behavior biometric map by mapping the original user payment behavior data to a knowledge graph, and uses the payment behavior biometric map to train the model of the transaction data to be paid to generate a transaction risk assessment prediction model. In this process, a large amount of data is needed to generate a relatively comprehensive payment behavior biometric map to show the association between different elements. Among the different elements, not only biometric information (for example, fingerprints) but also payment order information is involved. It should be noted that biometric information and order information belong to different modalities of information, that is, in the process of constructing the payment behavior biometric map, it is necessary to deal with multimodal data conversion and semantic alignment problems, which will not only consume a lot of computing resources, but also the robustness and generalization ability of the model are difficult to guarantee.

[0046] In response to this technical problem, the technical concept of the present application is to use intelligent algorithms to learn the semantic risk information in the transaction payment data that has been annotated, and to characterize the transaction risk type of the transaction payment data by quantifying the similarity between the semantic information in the transaction payment data and the semantic risk information in the transaction payment data that has been annotated in the feature common space, thereby intelligently identifying the risk type of the transaction payment data. That is, the problem of training a transaction risk assessment prediction model with multimodal data to perform risk assessment prediction is converted into the problem of metric matching between single modal data to achieve risk quantification. In this way, the relatively good consistency of the data structure of the sample set can be fully utilized to improve the robustness and generalization ability of model training.

[0047] In order to solve the above problems, the present disclosure provides a mobile terminal security payment management method and system, which obtains a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, a set of third transaction payment data marked as high-risk orders, and transaction data to be paid, and quantifies the similarity between the semantic information in the transaction data to be paid and the semantic risk information in the transaction payment data marked by the data in the feature common space to characterize the transaction risk type of the transaction data to be paid, thereby intelligently identifying the risk type of the transaction data to be paid. In this way, the relatively good consistency of the data structure of the sample set can be fully utilized to improve the robustness and generalization ability of model training.

[0048] The specific embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0049] Figure 1 is a flow chart of a mobile terminal secure payment management method according to an exemplary embodiment. Figure 1 As shown, the method includes:

[0050] Step 101: Acquire a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, and a set of third transaction payment data marked as high-risk orders;

[0051] Step 102: performing transaction payment data semantic feature abstract characterization on the set of the first transaction payment data marked as low-risk orders, the set of the second transaction payment data marked as medium-risk orders, and the set of the third transaction payment data marked as high-risk orders to obtain a low-risk order transaction payment data prototype characterization feature vector, a medium-risk order transaction payment data prototype characterization feature vector, and a high-risk order transaction payment data prototype characterization feature vector;

[0052] Step 103: Obtain transaction data to be paid;

[0053] Step 104: determine the risk type label of the transaction data to be paid based on the similarity between the transaction payment semantic feature information of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data, and the prototype characterization feature vector of the high-risk order transaction payment data;

[0054] Step 105: Encrypt data based on the risk type label of the transaction data to be paid.

[0055] Based on this, in the technical solution of the present application, first, a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, and a set of third transaction payment data marked as high-risk orders are obtained. Here, by obtaining a set of transaction payment data marked as orders of different risk levels, it is possible to ensure that the reference samples are diverse. That is, the reference samples cover transactions at different risk levels, such as low, medium, and high, which helps the model learn the semantic features and transaction rules of orders under various risk situations.

[0056] In one embodiment of the present disclosure, the set of first transaction payment data marked as low-risk orders, the set of second transaction payment data marked as medium-risk orders, and the set of third transaction payment data marked as high-risk orders are subjected to transaction payment data semantic feature abstract characterization to obtain low-risk order transaction payment data prototype characterization feature vectors, medium-risk order transaction payment data prototype characterization feature vectors, and high-risk order transaction payment data prototype characterization feature vectors, including: semantically encoding each first transaction payment data marked as low-risk order in the set of first transaction payment data marked as low-risk orders to obtain a set of low-risk order transaction payment data semantic encoding feature vectors; passing the set of low-risk order transaction payment data semantic encoding feature vectors through a gated switch prototype feature screening network to obtain the low-risk order transaction payment data prototype characterization feature vector; encoding the set of second transaction payment data marked as medium-risk orders and the set of third transaction payment data marked as high-risk orders in the same manner to obtain the medium-risk order transaction payment data prototype characterization feature vectors and the high-risk order transaction payment data prototype characterization feature vectors.

[0057] Then, semantic coding is performed on each of the first transaction payment data marked as low-risk orders in the set of the first transaction payment data marked as low-risk orders to obtain a set of semantic coding feature vectors of low-risk order transaction payment data. That is, the potential semantic information and transaction pattern contained in the set of the first transaction payment data marked as low-risk orders are extracted by semantic coding. Specifically, each of the first transaction payment data marked as low-risk orders may contain important information such as transaction content, information of both parties to the transaction, transaction environment and transaction behavior pattern. Among them, semantic coding can capture the semantic meaning of the transaction content, such as transaction type, transaction amount, transaction time and other information, thereby revealing the key transaction payment information of the order. In addition, the information of both parties to the transaction can provide key features of both parties to the transaction (buyer and seller), such as account information, transaction history, etc., which helps to understand the relationship and credit status between the transaction participants. The transaction environment includes information such as the transaction location, transaction equipment, and transaction network, which helps to judge the security and authenticity of the transaction. The transaction behavior pattern can reveal the behavior pattern in the transaction data, such as frequency, amount distribution and other characteristics, which helps to capture abnormal transaction behavior and fraud risks.

[0058] Considering that although obtaining a set of first transaction payment data marked as low-risk orders and semantically encoding them can capture the diverse feature information contained therein, the set of first transaction payment data marked as low-risk orders itself may contain some redundant information or background information, which will lead to a deviation in the expression of the core connotation of the transaction payment characteristics of low-risk orders. For example, in the transaction process of low-risk orders, the transaction location and transaction method often become the focus of risk assessment and identification information, while the transaction time may not be critical information for the results of risk assessment. In the technical solution of the present application, it is expected that the set of semantically encoded feature vectors of low-risk order transaction payment data will be passed through a gated switch prototype feature screening network to capture the core semantic expression and essential transaction behavior pattern of the set of semantically encoded feature vectors of low-risk order transaction payment data, thereby obtaining a prototype representation feature vector of low-risk order transaction payment data.

[0059] In one embodiment of the present disclosure, the set of semantically encoded feature vectors of the low-risk order transaction payment data is passed through a gated switch prototype feature screening network to obtain the prototype representation feature vector of the low-risk order transaction payment data, including: processing the set of semantically encoded feature vectors of the low-risk order transaction payment data with the following gated switch prototype feature screening formula to obtain the prototype representation feature vector of the low-risk order transaction payment data; wherein the gated switch prototype feature screening formula is:

[0060]

[0061]

[0062]

[0063] Among them, v i is the i-th low-risk order transaction payment data semantic encoding feature vector in the set of low-risk order transaction payment data semantic encoding feature vectors, v j is the jth low-risk order transaction payment data semantic encoding feature vector in the set of low-risk order transaction payment data semantic encoding feature vectors, ||·||1 represents the 1-norm of the feature vector, M is the length of the set of low-risk order transaction payment data semantic encoding feature vectors -1, V k The representation of the set of semantically encoded feature vectors of the low-risk order transaction payment data, D i represents the feature difference coefficient, mask represents masking processing, represents the feature difference coefficient after masking, exp(·) represents the natural exponential function operation, N s represents the total number of feature difference coefficients after masking, P kA feature vector is provided for representing a prototype of payment data for the low-risk order transaction.

[0064] Specifically, the gated switch prototype feature screening network highlights its own contribution and importance by measuring the overall difference information between each of the low-risk order transaction payment data semantic encoding feature vectors and other low-risk order transaction payment data semantic encoding feature vectors, and combines the gating mechanism to screen and distill such contribution and importance, thereby automatically selecting and emphasizing features that have a greater contribution to the target task, reducing the interference of redundant information and unimportant background information, and suppressing the deviation of feature distribution.

[0065] Further, in one embodiment of the present disclosure, the set of second transaction payment data marked as medium-risk orders and the set of third transaction payment data marked as high-risk orders are encoded in the same manner to obtain the medium-risk order transaction payment data prototype characterization feature vector and the high-risk order transaction payment data prototype characterization feature vector, including: semantically encoding each second transaction payment data marked as medium-risk order in the set of second transaction payment data marked as medium-risk orders to obtain a set of medium-risk order transaction payment data semantic encoding feature vectors; passing the set of medium-risk order transaction payment data semantic encoding feature vectors through the gated switch prototype feature screening network to obtain the medium-risk order transaction payment data prototype characterization feature vector; semantically encoding each third transaction payment data marked as high-risk order in the set of third transaction payment data marked as high-risk orders to obtain a set of high-risk order transaction payment data semantic encoding feature vectors; passing the set of high-risk order transaction payment data semantic encoding feature vectors through the gated switch prototype feature screening network to obtain the high-risk order transaction payment data prototype characterization feature vector.

[0066] In the technical solution of the present application, the set of the second transaction payment data marked as medium-risk orders and the set of the third transaction payment data marked as high-risk orders are encoded in the same way to extract and capture the transaction payment patterns of medium-risk orders and the transaction payment patterns of high-risk orders contained in the set of the second transaction payment data marked as medium-risk orders and the set of the third transaction payment data marked as high-risk orders, thereby obtaining the prototype characterization feature vector of the medium-risk order transaction payment data and the prototype characterization feature vector of the high-risk order transaction payment data.

[0067] At the same time, the transaction data to be paid is obtained, and the transaction data to be paid is semantically encoded to obtain a semantic encoding feature vector of the transaction data to be paid. That is, the rich semantic information contained in the transaction data to be paid and the transaction payment mode of the transaction behavior to be identified are extracted and captured in the same manner through the processing method of semantic encoding.

[0068] Subsequently, the hash similarity between the semantic coding feature vector of the transaction data to be paid and the prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the medium-risk order transaction payment data, and the prototype representation feature vector of the high-risk order transaction payment data is calculated to obtain a first hash similarity, a second hash similarity, and a third hash similarity. Here, by calculating the hash similarity, the degree of similarity between the semantic coding information of the transaction data to be paid and the abstract representation of the transaction risk prototype of orders of different risk levels can be quantified. This helps to judge the correlation and similarity between the data to be paid and orders of different risk levels. It should be understood that if there is a high hash similarity between the transaction data to be paid and the abstract representation of the transaction risk prototype under a certain risk level order, it may indicate that the transaction data to be paid has a high degree of match with the transaction pattern of the order of that risk level.

[0069] In one embodiment of the present disclosure, the risk type label of the transaction data to be paid is determined based on the degree of similarity between the transaction payment semantic feature information of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data and the prototype characterization feature vector of the high-risk order transaction payment data, including: semantically encoding the transaction data to be paid to obtain the semantic encoding feature vector of the transaction data to be paid; calculating the hash similarity between the semantic encoding feature vector of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data and the prototype characterization feature vector of the high-risk order transaction payment data to obtain a first hash similarity, a second hash similarity and a third hash similarity; based on the first hash similarity, the second hash similarity and the third hash similarity, determining the risk type label of the transaction data to be paid.

[0070] Further, the hash similarity between the semantic encoding feature vector of the transaction data to be paid and the prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the medium-risk order transaction payment data and the prototype representation feature vector of the high-risk order transaction payment data is calculated to obtain a first hash similarity, a second hash similarity and a third hash similarity, including: using a hash mapping function to map the semantic encoding feature vector of the transaction data to be paid, the prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the medium-risk order transaction payment data and the prototype representation feature vector of the high-risk order transaction payment data into a semantic encoding hash code of the transaction data to be paid, a prototype representation feature vector of the low-risk order transaction payment data, a prototype representation feature vector of the medium-risk order transaction payment data and a prototype representation feature vector of the high-risk order transaction payment data respectively. characterizing hash codes, medium-risk order transaction payment data prototype characterizing hash codes and high-risk order transaction payment data prototype characterizing hash codes; using a cosine similarity function to calculate the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype characterizing hash code of the low-risk order transaction payment data to obtain the first hash similarity; using a cosine similarity function to calculate the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype characterizing hash code of the medium-risk order transaction payment data to obtain the second hash similarity; using a cosine similarity function to calculate the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype characterizing hash code of the high-risk order transaction payment data to obtain the third hash similarity.

[0071] Further, based on the first hash similarity, the second hash similarity and the third hash similarity, the risk type label of the transaction data to be paid is determined; and based on the risk type label of the transaction data to be paid, the encryption method of the transaction data to be paid is determined. In an embodiment of the present application, the corresponding encryption method is determined according to the risk type label of the transaction to be paid. For example, a symmetric encryption algorithm, such as AES-256, is used for low-risk orders; an asymmetric encryption algorithm, such as RSA, is used for medium-risk orders; and a hybrid encryption method is used for high-risk orders, that is, a combination of symmetric and asymmetric encryption methods. In this way, the transaction data to be paid is encrypted based on the encryption method.

[0072] In one embodiment of the present disclosure, data encryption is performed based on the risk type label of the transaction data to be paid, including: determining an encryption method for the transaction data to be paid based on the risk type label of the transaction data to be paid; and encrypting the transaction data to be paid based on the encryption method.

[0073] In one embodiment of the present disclosure, the mobile terminal secure payment management method further includes a training step: training the gate switch prototype feature screening network. The training step includes: obtaining training data, the training data including a set of training first transaction payment data marked as low-risk orders, a set of training second transaction payment data marked as medium-risk orders, a set of training third transaction payment data marked as high-risk orders, and a true value of a first hash similarity, a true value of a second hash similarity, and a true value of a third hash similarity; semantically encoding each training first transaction payment data marked as a low-risk order in the set of training first transaction payment data marked as low-risk orders to obtain a set of semantic encoding feature vectors of training low-risk order transaction payment data; passing the set of semantic encoding feature vectors of training low-risk order transaction payment data through the gated switch prototype feature screening network to obtain a prototype representation feature vector of training low-risk order transaction payment data; encoding the set of training second transaction payment data marked as medium-risk orders and the set of training third transaction payment data marked as high-risk orders in the same way to obtain a prototype representation feature vector of training medium-risk order transaction payment data and a prototype representation feature vector of training high-risk order transaction payment data; obtaining training transaction data to be paid; The training transaction data to be paid is semantically encoded to obtain a semantic encoding feature vector of the training transaction data to be paid; the training low-risk order transaction payment data prototype representation feature vector is feature distribution optimized to obtain an optimized low-risk order transaction payment data prototype representation feature vector; the hash similarity between the semantic encoding feature vector of the training transaction data to be paid and the optimized low-risk order transaction payment data prototype representation feature vector, the training medium-risk order transaction payment data prototype representation feature vector and the training high-risk order transaction payment data prototype representation feature vector is calculated to obtain a training first hash similarity, a training second hash similarity and a training third hash similarity; the cross entropy values ​​between the training first hash similarity and the true value of the first hash similarity, between the training second hash similarity and the true value of the second hash similarity, and between the training third hash similarity and the true value of the third hash similarity are calculated respectively to obtain a first cross entropy value, a second cross entropy value and a third cross entropy value; the weighted sum of the first cross entropy value, the second cross entropy value and the third cross entropy value is used as the loss function value to train the gate switch prototype feature screening network.

[0074] In the above technical scheme, the set of semantic coding feature vectors of the training low-risk order transaction payment data expresses the coding semantic features of each training first transaction payment data marked as low-risk orders in the set of training first transaction payment data marked as low-risk orders. In this way, after the set of semantic coding feature vectors of the training low-risk order transaction payment data passes through the gated switch prototype feature screening network, the coding semantic features will be screened based on the coding semantic features of each training first transaction payment data marked as low-risk orders in the local sample space domain and the coding semantic prototype correlation of the global sample space domain of the set of training first transaction payment data marked as low-risk orders. While strengthening the significance of the coding semantic feature representation in the local sample space domain, the training low-risk order transaction payment data prototype representation feature vector will also show the discreteness of the feature distribution mapping relative to the local sample space domain due to the different weight reinforcement between the significant coding semantic features and the non-significant coding semantic features based on the attention mechanism.

[0075] In this way, when calculating the training first hash similarity between the semantic encoding feature vector of the training transaction data to be paid and the prototype representation feature vector of the training low-risk order transaction payment data, considering that the semantic encoding feature vector of the training transaction data to be paid expresses the encoded semantic features of the training transaction data to be paid based on the local sample space domain, the training prototype representation feature vector of the low-risk order transaction payment data will affect the calculation accuracy of the training first hash similarity due to the discreteness of the feature distribution mapping relative to the local sample space domain.

[0076] Based on this, the applicant of the present application optimizes the training low-risk order transaction payment data prototype characterization feature vector during the model training process, that is, when the loss function between the predicted training first hash similarity and the real first hash similarity is back-propagated through the training low-risk order transaction payment data prototype characterization feature vector, which is expressed as: the training low-risk order transaction payment data prototype characterization feature vector is characterized by feature distribution optimization using the following optimization formula to obtain the optimized low-risk order transaction payment data prototype characterization feature vector; wherein, the optimization formula is:

[0077]

[0078]

[0079] Among them, v i and v jare the eigenvalues ​​of the i-th and j-th positions of the training low-risk order transaction payment data prototype characterization feature vector V, V is the training low-risk order transaction payment data prototype characterization feature vector, V' is the optimized low-risk order transaction payment data prototype characterization feature vector, represents positional addition, Represents matrix multiplication.

[0080] That is, by introducing the local statistical information distribution of the training low-risk order transaction payment data prototype characterization feature vector V as an external information source to perform feature vector retrieval enhancement, so as to avoid the distribution illusion of the training low-risk order transaction payment data prototype characterization feature vector V caused by local overflow information distribution based on local statistical dense information structuring, thereby obtaining the information trusted response reasoning of the training low-risk order transaction payment data prototype characterization feature vector V based on the local distribution group dimension retention, so as to obtain the trusted distribution response of the training low-risk order transaction payment data prototype characterization feature vector V in the probability density space based on the discretized local feature distribution, thereby improving the feature distribution mapping effect of the training low-risk order transaction payment data prototype characterization feature vector facing the local sample space domain, so as to improve the calculation accuracy of the first hash similarity.

[0081] In summary, the above scheme is adopted to learn the semantic risk information in the transaction payment data that has been annotated by the data through the use of intelligent algorithms, and to characterize the transaction risk type of the transaction payment data by quantifying the similarity between the semantic information in the transaction payment data and the semantic risk information in the transaction payment data that has been annotated by the data in the feature common space, thereby intelligently identifying the risk type of the transaction payment data. That is, the problem of training a transaction risk assessment prediction model with multimodal data to perform risk assessment prediction is transformed into the problem of metric matching between single modal data to achieve risk quantification. In this way, the relatively good consistency of the data structure of the sample set can be fully utilized to improve the robustness and generalization ability of model training.

[0082] Figure 2 FIG. 1 is a block diagram of a mobile terminal secure payment management system according to an exemplary embodiment. Figure 2 As shown, the system 200 includes:

[0083] The payment data acquisition module 210 is used to acquire a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, and a set of third transaction payment data marked as high-risk orders;

[0084] The transaction payment data semantic feature abstract characterization module 202 is used to perform transaction payment data semantic feature abstract characterization on the set of the first transaction payment data marked as low-risk orders, the set of the second transaction payment data marked as medium-risk orders, and the set of the third transaction payment data marked as high-risk orders to obtain a low-risk order transaction payment data prototype characterization feature vector, a medium-risk order transaction payment data prototype characterization feature vector, and a high-risk order transaction payment data prototype characterization feature vector;

[0085] The transaction data to be paid is obtained by the module 203, which is used to obtain the transaction data to be paid;

[0086] A similarity determination module 204 is used to determine the risk type label of the transaction data to be paid based on the similarity between the transaction payment semantic feature information of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data, and the prototype characterization feature vector of the high-risk order transaction payment data;

[0087] The data encryption judgment module 205 is used to encrypt the data based on the risk type label of the transaction data to be paid.

[0088] Reference below Figure 3 , which shows a schematic diagram of the structure of an electronic device 600 suitable for implementing the embodiment of the present disclosure. The terminal device in the embodiment of the present disclosure may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0089] like Figure 3 As shown, the electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0090] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 3 The electronic device 600 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0091] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0092] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0093] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0094] The computer-readable medium may be included in the electronic device, or may exist independently without being installed in the electronic device.

[0095] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0096] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0097] The modules involved in the embodiments described in the present disclosure may be implemented by software or hardware. The name of a module does not limit the module itself in some cases. For example, a test parameter acquisition module may also be described as a "module for acquiring device test parameters corresponding to a target device".

[0098] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0099] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0100] Figure 4 FIG. 1 is an application scenario diagram of a mobile terminal secure payment management method according to an exemplary embodiment. Figure 4 As shown, in this application scenario, first, a set of first transaction payment data marked as low-risk orders is obtained (for example, Figure 4 C1 shown in FIG), a set of second transaction payment data marked as medium-risk orders (for example, Figure 4 C2) and a collection of third transaction payment data marked as high-risk orders (for example, Figure 4 C3 as shown in the figure); obtain the transaction data to be paid (for example, Figure 4 Then, the obtained set of first transaction payment data, the set of second transaction payment data, the set of third transaction payment data and the transaction to-be-paid data are input into a server (for example, Figure 4 In S) as shown in , the server is capable of processing the set of the first transaction payment data, the set of the second transaction payment data, the set of the third transaction payment data and the transaction data to be paid based on a mobile terminal security payment management algorithm to encrypt data based on a risk type label of the transaction data to be paid.

[0101] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

[0102] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0103] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims. Regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.

Claims

1. A mobile terminal secure payment management method, characterized in that: include: Acquire a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, and a set of third transaction payment data marked as high-risk orders; Performing transaction payment data semantic feature abstract characterization on the set of first transaction payment data marked as low-risk orders, the set of second transaction payment data marked as medium-risk orders, and the set of third transaction payment data marked as high-risk orders to obtain a low-risk order transaction payment data prototype characterization feature vector, a medium-risk order transaction payment data prototype characterization feature vector, and a high-risk order transaction payment data prototype characterization feature vector; Get the transaction data to be paid; Determine the risk type label of the transaction data to be paid based on the similarity between the transaction payment semantic feature information of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data, and the prototype characterization feature vector of the high-risk order transaction payment data; Encrypting data based on the risk type label of the transaction data to be paid; The set of the first transaction payment data marked as low-risk orders, the set of the second transaction payment data marked as medium-risk orders, and the set of the third transaction payment data marked as high-risk orders are subjected to transaction payment data semantic feature abstract characterization to obtain a low-risk order transaction payment data prototype characterization feature vector, a medium-risk order transaction payment data prototype characterization feature vector, and a high-risk order transaction payment data prototype characterization feature vector, including: Performing semantic coding on each of the first transaction payment data marked as low-risk orders in the set of first transaction payment data marked as low-risk orders to obtain a set of semantic coding feature vectors of low-risk order transaction payment data; Passing the set of semantically encoded feature vectors of the low-risk order transaction payment data through a gated switch prototype feature screening network to obtain the prototype representation feature vector of the low-risk order transaction payment data; In the same way, the set of the second transaction payment data marked as medium-risk orders and the set of the third transaction payment data marked as high-risk orders are encoded to obtain the prototype characterization feature vector of the medium-risk order transaction payment data and the prototype characterization feature vector of the high-risk order transaction payment data; The set of semantically encoded feature vectors of the low-risk order transaction payment data is passed through a gated switch prototype feature screening network to obtain the prototype representation feature vector of the low-risk order transaction payment data, including: The set of semantic encoding feature vectors of the low-risk order transaction payment data is processed using the following gate switch prototype feature screening formula to obtain the low-risk order transaction payment data prototype representation feature vector; wherein the gate switch prototype feature screening formula is: Among them, v i is the i-th low-risk order transaction payment data semantic encoding feature vector in the set of low-risk order transaction payment data semantic encoding feature vectors, v j is the jth low-risk order transaction payment data semantic encoding feature vector in the set of low-risk order transaction payment data semantic encoding feature vectors, ||·||1 represents the 1-norm of the feature vector, M is the length of the set of low-risk order transaction payment data semantic encoding feature vectors -1, V k The representation of the set of semantically encoded feature vectors of the low-risk order transaction payment data, D i represents the feature difference coefficient, mask represents masking processing, represents the feature difference coefficient after masking, exp(·) represents the natural exponential function operation, N s represents the total number of feature difference coefficients after masking, P k A feature vector is provided for representing a prototype of payment data for the low-risk order transaction.

2. The mobile terminal secure payment management method according to claim 1, characterized in that: The set of the second transaction payment data marked as medium-risk orders and the set of the third transaction payment data marked as high-risk orders are encoded in the same manner to obtain the prototype characterization feature vector of the medium-risk order transaction payment data and the prototype characterization feature vector of the high-risk order transaction payment data, including: Performing semantic coding on each of the second transaction payment data marked as medium-risk orders in the set of second transaction payment data marked as medium-risk orders to obtain a set of semantic coding feature vectors of medium-risk order transaction payment data; Passing the set of semantically encoded feature vectors of the medium-risk order transaction payment data through the gated switch prototype feature screening network to obtain the prototype representation feature vector of the medium-risk order transaction payment data; Performing semantic coding on each of the third transaction payment data marked as high-risk orders in the set of third transaction payment data marked as high-risk orders to obtain a set of semantic coding feature vectors of high-risk order transaction payment data; The set of semantically encoded feature vectors of the high-risk order transaction payment data is passed through the gated switch prototype feature screening network to obtain the prototype representation feature vector of the high-risk order transaction payment data.

3. The mobile terminal secure payment management method according to claim 2, characterized in that: Determining the risk type label of the transaction data to be paid based on the similarity between the transaction payment semantic feature information of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data, and the prototype characterization feature vector of the high-risk order transaction payment data, includes: Performing semantic coding on the transaction data to be paid to obtain a semantic coding feature vector of the transaction data to be paid; Calculating hash similarities between the semantic encoding feature vector of the transaction data to be paid and the prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the medium-risk order transaction payment data, and the prototype representation feature vector of the high-risk order transaction payment data to obtain a first hash similarity, a second hash similarity, and a third hash similarity; Based on the first hash similarity, the second hash similarity and the third hash similarity, a risk type label of the transaction data to be paid is determined.

4. The mobile terminal secure payment management method according to claim 3, characterized in that: Calculating the hash similarity between the semantic encoding feature vector of the transaction data to be paid and the prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the medium-risk order transaction payment data, and the prototype representation feature vector of the high-risk order transaction payment data to obtain a first hash similarity, a second hash similarity, and a third hash similarity, including: Use a hash mapping function to map the semantic encoding feature vector of the transaction data to be paid, the prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the medium-risk order transaction payment data, and the prototype representation feature vector of the high-risk order transaction payment data into a semantic encoding hash code of the transaction data to be paid, a prototype representation hash code of the low-risk order transaction payment data, a prototype representation hash code of the medium-risk order transaction payment data, and a prototype representation hash code of the high-risk order transaction payment data respectively; Using a cosine similarity function to calculate the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype representation hash code of the low-risk order transaction payment data to obtain the first hash similarity; Using a cosine similarity function to calculate the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype representation hash code of the medium-risk order transaction payment data to obtain the second hash similarity; A cosine similarity function is used to calculate the cosine similarity between the semantic encoding hash code of the transaction data to be paid and the prototype representation hash code of the high-risk order transaction payment data to obtain the third hash similarity.

5. The mobile terminal secure payment management method according to claim 4, characterized in that: Data encryption is performed based on the risk type label of the transaction data to be paid, including: Determining an encryption method for the transaction data to be paid based on the risk type label of the transaction data to be paid; Based on the encryption method, the transaction data to be paid is encrypted.

6. The mobile terminal secure payment management method according to claim 5, characterized in that: The method also includes a training step: training the gate-controlled switch prototype feature screening network.

7. The mobile terminal secure payment management method according to claim 6, characterized in that: The training step comprises: Acquire training data, wherein the training data includes a set of training first transaction payment data marked as low-risk orders, a set of training second transaction payment data marked as medium-risk orders, a set of training third transaction payment data marked as high-risk orders, and a true value of the first hash similarity, a true value of the second hash similarity, and a true value of the third hash similarity; Performing semantic encoding on each training first transaction payment data marked as a low-risk order in the set of training first transaction payment data marked as a low-risk order to obtain a set of semantic encoding feature vectors of training low-risk order transaction payment data; Passing the set of semantically encoded feature vectors of the training low-risk order transaction payment data through the gated switch prototype feature screening network to obtain a prototype representation feature vector of the training low-risk order transaction payment data; In the same way, the set of training second transaction payment data marked as medium-risk orders and the set of training third transaction payment data marked as high-risk orders are encoded to obtain a training medium-risk order transaction payment data prototype characterization feature vector and a training high-risk order transaction payment data prototype characterization feature vector; Get the training transaction payment data; Performing semantic encoding on the training transaction data to be paid to obtain a semantic encoding feature vector of the training transaction data to be paid; Performing feature distribution optimization on the training low-risk order transaction payment data prototype characterization feature vector to obtain an optimized low-risk order transaction payment data prototype characterization feature vector; Calculate the hash similarity between the semantic encoding feature vector of the training transaction data to be paid and the optimized prototype representation feature vector of the low-risk order transaction payment data, the prototype representation feature vector of the training medium-risk order transaction payment data, and the prototype representation feature vector of the training high-risk order transaction payment data to obtain a training first hash similarity, a training second hash similarity, and a training third hash similarity; Respectively calculating cross entropy values ​​between the training first Hash similarity and the true value of the first Hash similarity, between the training second Hash similarity and the true value of the second Hash similarity, and between the training third Hash similarity and the true value of the third Hash similarity to obtain a first cross entropy value, a second cross entropy value, and a third cross entropy value; The gate-controlled switch prototype feature screening network is trained by taking a weighted sum of the first cross-entropy value, the second cross-entropy value and the third cross-entropy value as a loss function value.

8. A mobile terminal security payment management system, characterized in that: include: A payment data acquisition module, used to acquire a set of first transaction payment data marked as low-risk orders, a set of second transaction payment data marked as medium-risk orders, and a set of third transaction payment data marked as high-risk orders; A transaction payment data semantic feature abstract characterization module is used to perform transaction payment data semantic feature abstract characterization on the set of first transaction payment data marked as low-risk orders, the set of second transaction payment data marked as medium-risk orders, and the set of third transaction payment data marked as high-risk orders to obtain a low-risk order transaction payment data prototype characterization feature vector, a medium-risk order transaction payment data prototype characterization feature vector, and a high-risk order transaction payment data prototype characterization feature vector; A transaction data acquisition module for obtaining transaction data to be paid, used to obtain transaction data to be paid; A similarity determination module, configured to determine a risk type label of the transaction data to be paid based on the similarity between the transaction payment semantic feature information of the transaction data to be paid and the prototype characterization feature vector of the low-risk order transaction payment data, the prototype characterization feature vector of the medium-risk order transaction payment data, and the prototype characterization feature vector of the high-risk order transaction payment data; A data encryption judgment module, used for encrypting data based on the risk type label of the transaction data to be paid; The transaction payment data semantic feature abstract representation module is used to: Performing semantic coding on each of the first transaction payment data marked as low-risk orders in the set of first transaction payment data marked as low-risk orders to obtain a set of semantic coding feature vectors of low-risk order transaction payment data; Passing the set of semantically encoded feature vectors of the low-risk order transaction payment data through a gated switch prototype feature screening network to obtain the prototype representation feature vector of the low-risk order transaction payment data; In the same way, the set of the second transaction payment data marked as medium-risk orders and the set of the third transaction payment data marked as high-risk orders are encoded to obtain the prototype characterization feature vector of the medium-risk order transaction payment data and the prototype characterization feature vector of the high-risk order transaction payment data; The set of semantic encoding feature vectors of the low-risk order transaction payment data is processed using the following gate switch prototype feature screening formula to obtain the low-risk order transaction payment data prototype representation feature vector; wherein the gate switch prototype feature screening formula is: Among them, v i is the i-th low-risk order transaction payment data semantic encoding feature vector in the set of low-risk order transaction payment data semantic encoding feature vectors, v j is the jth low-risk order transaction payment data semantic encoding feature vector in the set of low-risk order transaction payment data semantic encoding feature vectors, ||·||1 represents the 1-norm of the feature vector, M is the length of the set of low-risk order transaction payment data semantic encoding feature vectors -1, V k The representation of the set of semantically encoded feature vectors of the low-risk order transaction payment data, D i represents the feature difference coefficient, mask represents masking processing, represents the feature difference coefficient after masking, exp(·) represents the natural exponential function operation, N s represents the total number of feature difference coefficients after masking, P k A feature vector is provided for representing a prototype of payment data for the low-risk order transaction.

Citation Information

Patent Citations

  • Mobile terminal security payment management method and system

    CN117557270A

  • Transaction risk assessment method, apparatus and device, and computer readable storage medium

    CN111292085A

  • Risk control method and device for payment transaction

    CN114943538A

  • Payment data security protection method and device, electronic equipment and storage medium

    CN115001683A

  • Block chain-based payment transaction risk control method and device

    CN115018504A