Digital product secure transaction system

Abstract

The invention relates to the technical field of digital transaction security, and discloses a digital product security transaction system, which comprises a feature extraction module, a risk calculation module, a dynamic verification module, a list generation module and a data updating module. Compared with the prior art that transaction risk assessment is carried out only through a static rule, and particularly under the complex conditions of multiple devices, cross-regions, multiple amount types and the like, abnormal transaction behaviors cannot be effectively identified and user requirements cannot be considered, due to introduction of dynamic risk verification, multi-dimensional feature fusion and a security recommendation mechanism, the transaction risk assessment method and device have the advantages that the transaction risk assessment efficiency is improved, and the user experience is improved. Accurate risk assessment, safe recommendation and efficient transaction verification are realized, so that the problems of risk misjudgment and transaction fraud caused by model staticization are avoided, and the security of digital product transaction is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of digital transaction security, and in particular relates to a digital product security transaction system. Background Art

[0002] At present, with the rapid development of the digital economy, the scale of online transactions of digital products is expanding, and users' demands for convenience and diversification are increasing. However, the security issues of digital product transactions are becoming increasingly prominent. Traditional transaction systems mainly rely on static rules and a single-dimensional risk assessment mechanism, which is obviously insufficient in the face of modern complex transaction scenarios. For example, users may use multiple devices at the same time, such as mobile phones, tablets, and computers to conduct transactions. The traditional system has limited recognition capabilities for device fingerprints, and it is difficult to distinguish between normal multi-device logins and potential device forgeries; users often initiate transactions in different locations, but the existing technology lacks the ability to dynamically analyze the degree of deviation from the geographical location, and it is easy to misjudge legitimate cross-regional transactions as high-risk, or fail to detect real transaction fraud in a timely manner; the recommendation system is separated from the security mechanism: the traditional recommendation system only generates a recommendation list based on user preferences, without considering the security of the recommendation results, which may lead to frequent recommendations of high-risk products, increasing transaction security risks; safe transactions are the core goal of digital product transactions, and the accuracy and rationality of recommendations are actually to ensure the transaction security of users; recommended products not only need to meet user needs, but also consider their risk level, and give priority to providing users with products with high security levels, fundamentally reducing the risks in the transaction process. Therefore, there is an urgent need for a system that can dynamically assess risks and optimize the recommendation list in a complex transaction environment by combining the user's historical behavior characteristics and real-time transaction characteristics. The system should be able to accurately identify potential high-risk transaction behaviors under dynamic and changing transaction conditions, effectively adjust the recommendation results, give priority to recommending digital products with high security levels, dynamically adjust transaction verification rules through real-time risk scores, ensure transaction security, and balance user experience and transaction security. Summary of the invention

[0003] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to propose a digital product security transaction system, which aims to solve the technical problem that the prior art only uses static rules to evaluate transaction risks, especially under complex conditions such as multiple devices, cross-regions, and multiple amount types, and cannot effectively identify abnormal transaction behaviors and take into account user needs.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solution: The present invention provides a digital product security transaction system, comprising:

[0005] The feature extraction module obtains the user's historical behavior features from the platform database and generates the initial risk feature vector R0 based on the user's historical behavior features. The user's historical behavior features include the transaction frequency feature f t , transaction amount feature m t , device type characteristics d t , Geographical location characteristics t , transaction time characteristics t t ;

[0006] Among them, t is the preset time window, and the transaction frequency feature f t is the average number of transactions of the user within the preset time window t, and the transaction amount feature m t is the weighted average of the single transaction amount recorded by the user in historical transactions, and the device type feature d t The transaction device mainly used by the user is extracted based on the device fingerprint information, and the geographical location feature l t To determine the user's main transaction area through the IP address or GPS information at the time of transaction, the transaction time feature t t The time when high-frequency trading occurs;

[0007] The risk calculation module obtains real-time transaction features and calculates the real-time risk score R based on the initial risk feature vector R0 and the real-time transaction features. t ; Among them, the real-time transaction features include: current transaction amount feature m c , Current transaction device features d c 、Current transaction location characteristics c 、Current transaction frequency characteristics f c ;

[0008] The dynamic verification module, based on the real-time risk score R t , using the current transaction amount feature m c , Current transaction device features d c 、Current transaction location characteristics c 、Current transaction frequency characteristics f c Perform dynamic transaction verification, including:

[0009] Introducing the risk magnification factor α of transaction amount deviation m , according to the current transaction amount feature m c and transaction amount feature m t The difference between the transaction amount and the risk amplification factor α is used to calculate the deviation degree Δm of the transaction amount. If Δm is greater than the preset transaction amount threshold, the risk amplification factor α of the transaction amount deviation is used to calculate the deviation degree Δm of the transaction amount. m and real-time risk score R t Calculate the adjusted risk score P t; If Δm is not greater than the preset transaction amount threshold, the adjusted risk score P t With real-time risk score R t Stay consistent;

[0010] The risk weight coefficient β of introducing equipment abnormality d , compare the current transaction device characteristics d c and device type characteristics d t If the equipment is inconsistent, the secondary adjusted risk score P is calculated. t2 =P t β d If the equipment is consistent, the risk score after secondary adjustment is P t2 and risk score P t Stay consistent;

[0011] Introducing the risk magnification factor γ of position deviation l , calculate the final risk score P = P t γ l ;

[0012] Set the preset risk score threshold P th , when the final risk score P is greater than the preset risk score threshold P th When the final risk score P is not greater than the preset risk score threshold P th , the transaction verification passes;

[0013] List generation module: after the transaction verification is passed, the security recommendation score is calculated based on the improved weighted SlopeOne algorithm and a security recommendation list is generated;

[0014] The data update module sets up an encrypted payment channel for digital products in the security recommendation list. After the user selects and completes the payment, the user's historical behavior characteristics are updated in real time.

[0015] Preferably, in the feature extraction module, the initial risk feature vector R0 is generated using the formula:

[0016]

[0017] Among them, R0 is the initial risk feature vector, i is the index, x i is the i-th transaction feature in the user's historical behavior features, w i is the weight of the user’s i-th transaction feature, n is the length of the user’s historical behavior feature, and Normalize is the normalization function.

[0018] Preferably, in the risk calculation module, the real-time risk score R t The calculation formula is:

[0019] Rt =σ(W1R0+W2C t +b)

[0020] Among them, W1 and W2 are weight coefficients, b is the bias term, and C t is the real-time transaction feature, and σ is the activation function.

[0021] Preferably, in the list generation module, the steps of calculating the safety recommendation score based on the improved weighted SlopeOne algorithm and generating the safety recommendation list include:

[0022] Extract the user's historical digital product rating records from the platform database, compare the rating differences between the user's rated products and other digital products, and give priority to recommending digital products with small rating differences.

[0023] Preferably, in the list generation module, the step of calculating the safety recommendation score based on the improved weighted SlopeOne algorithm and generating the safety recommendation list further includes:

[0024] Assign priority weights to different products based on their security categories:

[0025] Give a higher recommendation weight to low-risk products;

[0026] Assign a lower recommendation weight to high-risk products.

[0027] Preferably, in the list generation module, the formula for calculating the safety recommendation score based on the improved weighted SlopeOne algorithm is:

[0028]

[0029] Among them, P j m is the user's security recommendation score for digital product j, i,j is the difference in recommendation scores between products i and j, n j is the security category weight of product j, and I is the set of digital products rated by users.

[0030] Preferably, in the data update module, the encrypted payment channel adopts the TLS protocol and combines the dynamic session key for data transmission.

[0031] The beneficial effect of the present invention is that compared with the prior art that only uses static rules to conduct transaction risk assessment, especially under complex conditions such as multiple devices, cross-regions, and multiple amounts, the technical problem of being unable to effectively identify abnormal transaction behaviors and take into account user needs is solved. Due to the introduction of dynamic risk verification, multi-dimensional feature fusion, and security recommendation mechanism, the present application realizes accurate risk assessment, security recommendation, and efficient transaction verification, thereby avoiding the risk misjudgment and transaction fraud caused by the static model, and improving the security of digital product transactions. . BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0033] Figure 1 This is a flow chart of a first embodiment of a digital product secure transaction system provided by the present invention.

[0034] Figure 2 A schematic diagram of a device of a digital product secure transaction system provided by the present invention. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] Embodiment 1: Figure 1 FIG. 1 is a flow chart of the first embodiment of the digital product security transaction system of the present invention, and provides the first embodiment of the digital product security transaction system of the present invention.

[0037] In a first embodiment, the digital product secure transaction system includes:

[0038] The feature extraction module obtains the user's historical behavior features from the platform database and generates the initial risk feature vector R0 based on the user's historical behavior features. The user's historical behavior features include the transaction frequency feature f t , transaction amount feature m t , device type characteristics d t , Geographical location characteristics t , transaction time characteristics t t ;

[0039] Among them, t is the preset time window, and the transaction frequency feature f t is the average number of transactions of the user within the preset time window t, and the transaction amount feature m t is the weighted average of the single transaction amount recorded by the user in historical transactions, and the device type feature d t The transaction device mainly used by the user is extracted based on the device fingerprint information, and the geographical location feature l t To determine the user's main transaction area through the IP address or GPS information at the time of transaction, the transaction time feature t t The time when high-frequency trading occurs;

[0040] It should be noted that the extraction and analysis of these features provide a solid foundation for subsequent risk assessment. The transaction frequency feature can be used to measure the activity of user transaction behavior; the transaction amount feature can be used to evaluate the stability and deviation of the user's transaction amount; the device type feature can be used to determine the credibility of the user's transaction device; the geographic location feature can be used to effectively identify abnormalities in the user's transaction location; and the transaction time feature can be used to capture the temporal regularity of the user's transaction behavior. The above features are interrelated and together constitute the user's historical behavior portrait.

[0041] It is understandable that the extraction and analysis of user historical behavior characteristics not only helps to evaluate the user's trading habits and risks, but also can dynamically adapt to complex trading environments. For example, when the user's historical transaction frequency is high and the transaction amount is relatively stable, the system can consider their behavior to be more credible. However, if a transaction amount suddenly deviates from the historical amount range or the device is inconsistent with the historical device, there may be a risk of abnormal transactions.

[0042] It should be understood that the feature extraction module supports the joint analysis of multi-dimensional features and can dynamically perceive the changing trends between features. For example, by combining the device type feature with the geographic location feature, it can be determined whether the user is trading in a trusted device and location; by combining the transaction amount feature with the transaction time feature, it can be determined whether the user is conducting abnormal transactions during non-high-frequency trading time periods.

[0043] For example, if a user's historical transaction records show that he or she mainly makes small top-ups between 6pm and 8pm every day, and the device is an iPhone, and the transaction location is within a fixed range in a certain city. However, if a transaction occurs at 2am, the device is changed to Android, and the transaction amount is significantly higher, the system will identify abnormal deviations of the above features through the feature extraction module, providing important basis for subsequent risk assessment.

[0044] Through the design of this feature extraction module, the system can efficiently build a user's historical behavior portrait, provide accurate data support for dynamic risk assessment and subsequent transaction verification, thereby significantly improving the security and accuracy of digital product transactions.

[0045] The risk calculation module obtains real-time transaction features and calculates the real-time risk score R based on the initial risk feature vector R0 and the real-time transaction features. t ; Among them, the real-time transaction features include: current transaction amount feature m c , Current transaction device features d c 、Current transaction location characteristics c 、Current transaction frequency characteristics f c ;

[0046] It should be noted that the risk calculation module realizes dynamic real-time risk score calculation by combining the user's initial risk feature vector and the above-mentioned real-time transaction features. The initial risk feature vector provides a stable portrait of the user's historical transaction behavior, while the real-time transaction features reflect the dynamic characteristics of the current transaction environment. The combination of the two can comprehensively evaluate the normality and potential risks of transaction behavior. The real-time risk score is obtained based on the dynamic adjustment of the initial risk score. It not only reflects the stability of historical behavior, but also reflects the environmental security of the current transaction.

[0047] It is understandable that the design of the risk calculation module can effectively capture abnormal deviations in user transaction behaviors. For example, by comparing the current transaction amount characteristics with the historical transaction amount characteristics, the degree of deviation of the transaction amount can be calculated; by detecting the consistency of the current transaction device characteristics with the historical device characteristics, the risk of device counterfeiting can be identified; by analyzing the geographic location characteristics, the legitimacy of the user's transaction location can be determined; by real-time monitoring of transaction frequency characteristics, high-frequency trading behaviors in a short period of time can be captured, thereby identifying potential abnormal transaction patterns.

[0048] It should be understood that the calculation of real-time risk scores is a dynamic adjustment process that combines the correlation between multiple features. For example, when the user's current transaction amount deviates significantly from the historical transaction amount and the device is inconsistent with the historical device, the system will dynamically increase the risk score to indicate potential transaction risks; when the user's transaction frequency is consistent with historical characteristics and the device and geographical location have not deviated, the risk score remains at a low level, indicating that the transaction environment is relatively safe.

[0049] For example, suppose the user's historical transaction behavior characteristics show that his commonly used device is iPhone, the transaction amount is between 100-500 yuan, the transaction location is mainly concentrated in a certain city, and the transaction frequency is once a day. However, a transaction occurred in the early morning, the device was changed to Android, the amount reached 2,000 yuan, and the geographical location deviated to another country. Through the risk calculation module, the system combines the initial risk feature vector and real-time transaction characteristics to identify three risk points: transaction amount deviation, device forgery, and abnormal geographical location, dynamically adjust the risk score, and finally judge the transaction as a high-risk transaction.

[0050] Through the design of the risk calculation module, the system can quickly respond to changes in the trading environment, evaluate potential transaction risks in real time, effectively prevent fraudulent transactions, and provide users with a safer digital product trading experience.

[0051] The dynamic verification module, based on the real-time risk score R t , using the current transaction amount feature m c , Current transaction device features d c 、Current transaction location characteristics c 、Current transaction frequency characteristics f c Perform dynamic transaction verification, including:

[0052] Introducing the risk magnification factor α of transaction amount deviation m , according to the current transaction amount feature m c and transaction amount feature m t The difference between the transaction amount and the risk amplification factor α is used to calculate the deviation degree Δm of the transaction amount. If Δm is greater than the preset transaction amount threshold, the risk amplification factor α of the transaction amount deviation is used to calculate the deviation degree Δm of the transaction amount. m and real-time risk score R t Calculate the adjusted risk score P t ; If Δm is not greater than the preset transaction amount threshold, the adjusted risk score P t With real-time risk score R t Stay consistent;

[0053] The risk weight coefficient β of introducing equipment abnormality d , compare the current transaction device characteristics d c and device type characteristics d t If the equipment is inconsistent, the secondary adjusted risk score P is calculated. t2 =P t β d If the equipment is consistent, the risk score after secondary adjustment is P t2 and risk score P t Stay consistent;

[0054] Introducing the risk magnification factor γ of position deviationl , calculate the final risk score P = P t γ l ;

[0055] Set the preset risk score threshold P th , when the final risk score P is greater than the preset risk score threshold P th When the final risk score P is not greater than the preset risk score threshold P th , the transaction verification passes;

[0056] It should be noted that the dynamic verification module realizes multi-level risk assessment through comprehensive analysis of the degree of deviation of transaction amount, device consistency and geographic location deviation. The introduction and adjustment of various risk factors enable the risk assessment model to dynamically adapt to the complex and changing transaction environment. For example, by calculating the degree of deviation of transaction amount, it is possible to identify abnormal deviation of amount in user transaction behavior; through device consistency verification, it is possible to determine the risk of device counterfeiting or unauthorized device; through dynamic analysis of geographic location, it is possible to capture potential transaction location anomalies.

[0057] It is understandable that the module takes into account the relative weight and influence of each feature when dynamically adjusting the risk score, thereby achieving a flexible and accurate transaction verification process. For example, the device consistency risk usually has a larger weight because device counterfeiting may directly lead to transaction security issues; while the deviation of the transaction amount and the risk of geographical location will dynamically adjust the weight according to the actual scenario.

[0058] It should be understood that the design of the dynamic verification module can effectively reduce the misjudgment rate and risk tolerance rate, while improving verification efficiency. For example, when the user's transaction amount is close to the historical record, the device is consistent and the location is not deviated, the system can quickly determine the transaction security; on the contrary, when multiple features are abnormal at the same time, the system will dynamically increase the risk score, thereby preventing potential high-risk transactions.

[0059] For example, suppose a user's historical transaction behavior shows that his commonly used device is iPhone, the transaction amount is between 500-1000 yuan, and the main transaction location is concentrated in downtown Shanghai. The transaction amount is 5000 yuan, the device is changed to Android, and the transaction location is a city abroad. The dynamic verification module will calculate the transaction amount deviation degree Δm, device consistency and location deviation degree based on the above characteristics. The final risk score P increases significantly, and the system will reject the transaction to ensure the security of the user's account.

[0060] Through the introduction of the dynamic verification module, the system can quickly respond to changes in the transaction environment, comprehensively analyze multi-dimensional features, significantly improve transaction security, effectively prevent fraud and abnormal behavior, and ensure the reliability of digital product transactions.

[0061] List generation module: after the transaction verification is passed, the security recommendation score is calculated based on the improved weighted SlopeOne algorithm and a security recommendation list is generated;

[0062] It should be noted that the list generation module generates a recommendation list that is both safe and relevant by combining user historical rating data, rating deviation and product risk weight. The improved weighted Slope One algorithm adds product safety weights to the traditional rating calculation, giving priority to recommending products with lower risks, thereby effectively reducing the possibility of transaction security issues caused by recommending high-risk products.

[0063] It is understandable that the list generation module not only focuses on user preferences, but also fully considers the security level and risk assessment results of digital products. For example, the user's preference for low-value products may be directly reflected in the score, while high-risk products may be lowered in recommendation priority due to their low security category weight. Through this design, the system is able to balance user needs and transaction security.

[0064] It should be understood that the recommendation logic of this module fully utilizes the results of user historical behavior and current transaction verification to ensure that the recommendation list greatly reduces security risks while meeting user needs. For example, products with high scores but high risks will be dynamically lowered in priority by the system, while products with high security levels will be recommended first, thus building a list that takes into account both accurate recommendations and security guarantees.

[0065] For example: A user has a rating record for the following digital products:

[0066] Product A (phone recharge): Rating 5 points;

[0067] Product B (traffic recharge): score 4 points;

[0068] Product C (virtual assets): score 3 points, security category weight is 0.8;

[0069] Product D (value-added service): No rating record, safety category weight is 1.2.

[0070] Based on user rating data and the improved weighted Slope One algorithm, the system calculates the recommendation score of unrated product D, and dynamically adjusts the recommendation priority based on its safety category weight. Ultimately, product D is recommended first because of its higher safety category weight, while product C is downgraded because of its lower weight, thus ensuring that the recommendation results are safe and reliable.

[0071] Through the list generation module, the system achieves accurate and safe digital product recommendations, which improves the user experience while significantly improving the security of the transaction process and reducing transaction problems that may be caused by recommending high-risk products.

[0072] The data update module sets up an encrypted payment channel for digital products in the security recommendation list. After the user selects and completes the payment, the user's historical behavior characteristics are updated in real time.

[0073] It should be noted that the data update module sets up an encrypted payment channel to ensure the security of transaction data during the user's payment process and prevent the data from being stolen or tampered with during transmission. At the same time, after the user completes the payment, the system will record the user's historical behavior characteristics in real time, including key characteristics such as transaction amount, transaction device, transaction location and transaction time. These data will be directly updated to the user's historical transaction records to optimize subsequent risk assessment and recommendation list generation.

[0074] It is understandable that the setting of encrypted payment channels is the basis of transaction security. By adopting dynamic key exchange and data encryption technology, the system can protect users' sensitive payment information, such as transaction amount, payment account information, etc.; real-time updating of user historical behavior characteristics is to ensure the accuracy of user portraits, so that the system can capture user behavior changes more timely, adjust risk assessment strategies and recommendation algorithms. For example, changes in users' transaction frequency or transaction amount habits can be reflected in their historical behavior data in real time, enhancing the system's ability to respond to abnormal transactions.

[0075] It should be understood that the data update module is not only a record of user behavior data, but also a dynamic optimization process. Each update of historical behavior characteristics will directly affect the user's initial risk feature vector, so that the system can make judgments based on more accurate user portraits in subsequent risk assessments and recommendation calculations. For example, as the user's device changes, the system can dynamically adjust the judgment logic for device consistency; as the user's transaction frequency changes, the system can re-evaluate its transaction activity, thereby optimizing the calculation of risk scores.

[0076] For example, a user selects a phone recharge product from the security recommendation list, the transaction amount is 100 yuan, the device is an iPhone, the transaction location is downtown Beijing, and the transaction time is 8 pm. The system protects the user's payment information through an encrypted payment channel. After the transaction is completed, the following data is updated in real time to the user's historical behavior record:

[0077] The transaction frequency feature updates the average number of transactions of the user in the last 30 days. If this is the user's fifth transaction this month, the transaction frequency is updated to 0.167 times / day.

[0078] The transaction amount feature updates the user's average transaction amount based on the weighted average formula. If the previous weighted average was 120 yuan, it will be updated to 116 yuan after this transaction.

[0079] The device type feature records that the device for this transaction is an iPhone and updates it to the device that the user primarily uses.

[0080] Geographical location characteristics: The transaction is recorded as taking place in downtown Beijing. If the previous main transaction location was in downtown Beijing, the location will remain consistent.

[0081] Trading time features: record the trading time as 8 pm and update the user's high-frequency trading time period.

[0082] Through real-time updates of this data, the system can timely optimize the user's initial risk feature vector and dynamically adjust the subsequent risk assessment logic and recommendation list generation strategy. For example, if the user later changes the device or the transaction amount deviates significantly from the historical average, the system can capture these changes more quickly to ensure the accuracy of risk assessment and transaction security.

[0083] In addition, the present invention also provides a digital product security transaction device, please refer to Figure 2, a digital product security transaction device includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a digital product security transaction system in the above-mentioned embodiment 1. A digital product security transaction device in the embodiment of the present invention may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. A digital product security transaction device is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention. A digital product security transaction device may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM) 1004. In RAM1004, various programs and data required for the operation of a digital product security transaction device are also stored. The processing device 1001, ROM1002 and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow a digital product security transaction device to communicate wirelessly or wired with other devices to exchange data. Although a digital product security transaction device with various systems is shown in the figure, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0084] The present invention also provides a computer program product, including a computer program, which implements the steps of the above-mentioned digital product security transaction system when executed by a processor. The computer program product provided by the present invention can solve the technical problem of digital product security transaction. Compared with the prior art, the beneficial effects of the computer program product provided by the present invention are the same as the beneficial effects of the digital product security transaction system provided by the above-mentioned embodiment, which will not be repeated here.

[0085] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the system of the embodiments disclosed in the present invention are executed.

[0086] It should be understood that the various parts disclosed in the present invention can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0087] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A digital product security transaction system, characterized in that: The system includes: The feature extraction module obtains the user's historical behavior features from the platform database and generates the initial risk feature vector R0 based on the user's historical behavior features. The user's historical behavior features include the transaction frequency feature f t , transaction amount feature m t , device type characteristics d t 、Geographical location features t , transaction time characteristics t t ; Among them, t is the preset time window, and the transaction frequency feature f t is the average number of transactions of the user within the preset time window t, and the transaction amount feature m t is the weighted average of the single transaction amount recorded by the user in historical transactions, and the device type feature d t The transaction device mainly used by the user is extracted based on the device fingerprint information, and the geographical location feature l t To determine the user's main transaction area through the IP address or GPS information at the time of transaction, the transaction time feature t t The time when high-frequency trading occurs; The risk calculation module obtains real-time transaction features and calculates the real-time risk score R based on the initial risk feature vector R0 and the real-time transaction features. t ; Among them, the real-time transaction features include: current transaction amount feature m c , Current transaction device features d c 、Current transaction location characteristics c 、Current transaction frequency characteristics f c ; The dynamic verification module, based on the real-time risk score R t , using the current transaction amount feature m c , Current transaction device features d c 、Current transaction location characteristics c 、Current transaction frequency characteristics f c Perform dynamic transaction verification, including: Introducing the risk magnification factor α of transaction amount deviation m , according to the current transaction amount feature m c and transaction amount feature m t The difference between the transaction amount and the risk amplification factor α is used to calculate the deviation degree Δm of the transaction amount. If Δm is greater than the preset transaction amount threshold, the risk amplification factor α of the transaction amount deviation is used to calculate the deviation degree Δm of the transaction amount. m and real-time risk score R t Calculate the adjusted risk score P t ; If Δm is not greater than the preset transaction amount threshold, the adjusted risk score P t With real-time risk score R t Stay consistent; The risk weight coefficient β of introducing equipment abnormality d , compare the current transaction device characteristics d c and device type characteristics d t If the equipment is inconsistent, the secondary adjusted risk score P is calculated. t2 =P t β d If the equipment is consistent, the risk score after secondary adjustment is P t2 and risk score P t Stay consistent; Introducing the risk magnification factor γ of position deviation l , calculate the final risk score P = P t γ l ; Set the preset risk score threshold P th , when the final risk score P is greater than the preset risk score threshold P th When the final risk score P is not greater than the preset risk score threshold P th , the transaction verification passes; List generation module: after the transaction verification is passed, the security recommendation score is calculated based on the improved weighted SlopeOne algorithm and a security recommendation list is generated; The data update module sets up an encrypted payment channel for digital products in the security recommendation list. After the user selects and completes the payment, the user's historical behavior characteristics are updated in real time.

2. A digital product secure transaction system as claimed in claim 1, characterized in that: In the feature extraction module, the initial risk feature vector R0 is generated using the formula: Among them, R0 is the initial risk feature vector, i is the index, x i is the i-th transaction feature in the user's historical behavior features, w i is the weight of the user’s i-th transaction feature, n is the length of the user’s historical behavior feature, and Normalize is the normalization function.

3. A digital product secure transaction system as claimed in claim 1, characterized in that: In the risk calculation module, the real-time risk score R t The calculation formula is: R t =σ(W1R0+W2C t +b) Among them, W1 and W2 are weight coefficients, b is the bias term, and C t is the real-time transaction feature, and σ is the activation function.

4. A digital product secure transaction system as claimed in claim 1, characterized in that: In the list generation module, the steps of calculating the safety recommendation score and generating the safety recommendation list based on the improved weighted SlopeOne algorithm include: Extract users' historical rating records of digital products from the platform database, compare the rating differences between the products rated by users and the ratings of other digital products, and give priority to recommending digital products with small rating differences.

5. A digital product secure transaction system as claimed in claim 1, characterized in that: In the list generation module, the steps of calculating the safety recommendation score and generating the safety recommendation list based on the improved weighted SlopeOne algorithm also include: Assign priority weights to different products based on their security categories: Give a higher recommendation weight to low-risk products; Assign a lower recommendation weight to high-risk products.

6. A digital product secure transaction system as claimed in claim 4, characterized in that: In the list generation module, the formula for calculating the safety recommendation score based on the improved weighted SlopeOne algorithm is: Among them, P j The security recommendation score for digital product j by the user, m i,j is the difference in recommendation scores between products i and j, n j is the security category weight of product j, and I is the set of digital products rated by users.

7. A digital product secure transaction system as claimed in claim 1, characterized in that: In the data update module, the encrypted payment channel uses the TLS protocol and combines dynamic session keys for data transmission.

Images