A multi-party secure transaction method based on cross-border e-commerce business rules

Abstract

This invention relates to the field of multi-party secure transaction technology, specifically disclosing a multi-party secure transaction method based on cross-border e-commerce business rules, including the following steps: Step S1: Create a transaction processing center, obtain its historical transaction data, and filter high-frequency e-commerce users and their authentication information; Step S2: Construct a time series of high-frequency authentication information, select time nodes to obtain text vectors to form a set, and use principal component analysis to obtain text change curves; Step S3: Determine whether the high-frequency authentication information is inertial change information based on the curve. If so, determine the stable interval and establish a pre-cached center to store the information within the latest stable interval; Step S4: When cross-border e-commerce users conduct transactions, obtain information extraction and transaction time periods. If inertial change information is involved and both are within the same stable interval, the transaction processing center obtains the information from the pre-cached center. This invention reduces redundant calculations and data interaction, improves the efficiency of cross-border e-commerce multi-party secure transactions, and reduces the computational burden.

Description

Technical Field

[0001] This invention relates to the field of multi-party secure transaction technology, and specifically to a multi-party secure transaction method based on cross-border e-commerce business rules. Background Technology

[0002] Within the existing technological framework, the security requirements for transaction computation in cross-border e-commerce are significantly higher than in conventional business scenarios. This characteristic stems from the unique nature of cross-border e-commerce: its transaction chain involves entities from multiple countries and regions, and all parties must strictly adhere to the data privacy regulations of different jurisdictions during data exchange. Simultaneously, they must guard against security threats such as cross-border data breaches, man-in-the-middle attacks, and fraudulent transactions. Therefore, the entire transaction computation process must be conducted under encryption protection, directly leading to an exponential increase in computational complexity.

[0003] Meanwhile, the user information authentication tasks in cross-border e-commerce further exacerbate the computational burden. Unlike local e-commerce, cross-border e-commerce users need to undergo multi-dimensional identity verification, which involves cross-comparison and joint calculation of multi-source data. For example, payment institutions need to collaborate with banks to verify users' cross-border payment permissions, logistics companies need to share user identity information with customs to complete customs clearance, and tax authorities need to calculate tariffs in real time based on transaction data. Such multi-entity, multi-dimensional authentication and calculation tasks lead to a surge in the amount of input data for multi-party secure computation (MPC) and require the handling of complex business logic, resulting in low efficiency of existing MPC technologies in cross-border e-commerce. Summary of the Invention

[0004] The purpose of this invention is to provide a multi-party secure transaction method based on cross-border e-commerce business rules, and to solve the following technical problems.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A multi-party secure transaction method based on cross-border e-commerce business rules includes the following steps:

[0007] Step S1: Establish a transaction processing center, obtain historical transaction data from the transaction processing center, and filter out high-frequency e-commerce users and their high-frequency authentication information based on the historical transaction data;

[0008] Step S2: Establish the time series of the high-frequency authentication information, select several time nodes in the time series, obtain the text vector of the high-frequency authentication information at each time node, and obtain the text vector set; obtain the text change curve of the high-frequency authentication information based on principal component analysis technology;

[0009] Step S3: Based on the text change curve, determine whether the high-frequency authentication information is inertly changing information; if the high-frequency authentication information is inertly changing information, determine the stable range of the inertly changing information; establish a pre-cache center, and store the inertly changing information in the latest stable range in the pre-cache center in real time;

[0010] Step S4: When a cross-border e-commerce user performs a transaction task through the transaction processing center, the information extraction time and transaction time period are obtained; when the transaction task involves lazy change information, if the information extraction time and the transaction time period are in the same stable range, the transaction processing center directly obtains the lazy change information from the pre-cached center.

[0011] As a further aspect of the present invention: the transaction processing center is used to execute transaction tasks between various cross-border e-commerce users, and the historical transaction data includes the transaction tasks of each cross-border e-commerce user, as well as the authentication information of the cross-border e-commerce users involved in each transaction task.

[0012] As a further aspect of the present invention: the process of acquiring high-frequency e-commerce users includes:

[0013] Obtain the total number of all transaction tasks in the historical transaction data, denoted as the total transaction volume, and obtain the number of transaction tasks participated in by cross-border e-commerce users in the historical transaction data, denoted as the user transaction volume. The participation rate of cross-border e-commerce users is Pop = n / N × 100%, where n is the user transaction volume and N is the total transaction volume.

[0014] If the participation rate of the cross-border e-commerce user is greater than or equal to a preset threshold, then the cross-border e-commerce user is a high-frequency e-commerce user, and the threshold is set within the range of [50%, 90%]; otherwise, the cross-border e-commerce user is not a high-frequency e-commerce user; and the authentication information of the high-frequency e-commerce user is recorded as high-frequency authentication information.

[0015] As a further aspect of the present invention: the process of establishing the time series of the high-frequency authentication information includes:

[0016] The time of the first upload of the high-frequency authentication information in the historical transaction data is obtained and recorded as the first upload time, and the time of the latest upload of the high-frequency authentication information is obtained and recorded as the latest upload time; the time period between the first upload time and the latest upload time is recorded as the time series of the high-frequency authentication information.

[0017] As a further aspect of the present invention: the process of obtaining the text change curve of the high-frequency authentication information includes:

[0018] Obtain a set of text vectors {(x1, y1), (x2, y2), ..., (xm, ym)}, where (xm, ym) represents the coordinates of the m-th text vector, and m is the total number of text vectors; based on the set of text vectors, obtain the mean of the set of text vectors along the x-axis. and the mean in the y-dimension Where xi represents the coordinate value of the i-th text vector in the x-dimensional direction, yi represents the coordinate value of the i-th text vector in the y-dimensional direction, i∈[1,m] and i is a positive integer;

[0019] And obtain the standard deviation of the text vector set in the x-dimensional dimension. and the standard deviation in the y-dimension

[0020] Then we obtain the standardized value of the i-th text vector in the x-dimensional direction. and standardized values ​​in the y-dimension Obtain the covariance matrix of the text vector set. The maximum eigenvalue of the text vector set is obtained based on the covariance matrix, and the eigenvector corresponding to the maximum eigenvalue is obtained.

[0021] The feature vector is denoted as (a, b). For any text vector (xi, yi), the projection value of the text vector (xi, yi) is z = a*xi′ + b*yi′. Based on the projection value of the high-frequency authentication information at each time node, the text change curve of the high-frequency authentication information is established.

[0022] As a further aspect of the present invention: the process of determining whether the high-frequency authentication information is inertly changing information includes:

[0023] Based on the text change curve, the volatility of the high-frequency authentication information is obtained. Based on the volatility, a volatility threshold is set. If the volatility of the text change curve is less than the volatility threshold, the high-frequency authentication information corresponding to the text change curve is inert change information; otherwise, the high-frequency authentication information corresponding to the text change curve is not inert change information.

[0024] As a further aspect of the present invention: the process of obtaining the volatility of the high-frequency authentication information includes:

[0025] Several time window values ​​are set, each time window value being a time period composed of several time nodes, with different time window values ​​containing different numbers of time nodes; based on the time window values, several time windows are extracted from the time series, and the curve segment corresponding to each time window on the text change curve is obtained;

[0026] Obtain the number of extreme points on the curve segment, excluding the two endpoints of the curve segment; based on the number of extreme points on all curve segments, obtain the average number of extreme points, which is denoted as the average number of extreme points corresponding to the time window value;

[0027] Based on the average number of extreme points corresponding to each time window value, a regression line is obtained by fitting using the least squares method, and the slope of the regression line is obtained and denoted as the volatility of the text change curve.

[0028] As a further aspect of the present invention: the process of determining the stable interval of the inertial change information includes:

[0029] Several reference points are selected equally on the text change curve, and the slope of the tangent line of the text change curve at each reference point is obtained; the slope difference K = |kw-k(w-1)| between the reference point and the previous reference point is obtained in real time, where kw represents the k-th reference point and k(w-1) represents the (k-1)-th reference point; if there are several consecutive reference points on the text change curve where the slope difference K is 0, the horizontal coordinate interval formed by these reference points is recorded as the stable interval.

[0030] The beneficial effects of this invention are:

[0031] Cross-border e-commerce transactions involve multi-party data interaction and encrypted computation. In traditional models, user authentication information must be repeatedly verified for each transaction, leading to a surge in computational load. This method focuses on core data in high-frequency transaction scenarios by filtering high-frequency e-commerce users and their authentication information, avoiding ineffective computation on low-frequency data. Furthermore, principal component analysis is used to construct text change curves, identifying inertial change information with relatively small fluctuations, and pre-storing data within its stable range in a pre-cached center. When a user transacts, if the information retrieval time and the transaction time period are within the same stable range, the pre-cached data can be directly accessed, reducing the frequency of real-time encrypted computation and multi-source data interaction, thereby reducing computational complexity and improving transaction processing efficiency. This invention systematically solves the problems of low efficiency and heavy computational burden in multi-party secure computation in cross-border e-commerce by combining data feature analysis, a pre-caching mechanism, and a dynamic adaptation strategy, providing technical support for improving the stability and scalability of cross-border transactions. Attached Figure Description

[0032] The invention will now be further described with reference to the accompanying drawings.

[0033] Figure 1 This is a schematic diagram of the structure of a multi-party secure transaction method based on cross-border e-commerce business rules according to the present invention. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] Please see Figure 1 As shown, this invention is a multi-party secure transaction method based on cross-border e-commerce business rules, comprising the following steps:

[0036] Step S1: Establish a transaction processing center, which is used to execute transaction tasks between various cross-border e-commerce users; obtain historical transaction data from the transaction processing center, which includes transaction tasks of various cross-border e-commerce users and authentication information of cross-border e-commerce users involved in each transaction task; based on the historical transaction data, filter out high-frequency e-commerce users and record the authentication information of the high-frequency e-commerce users as high-frequency authentication information.

[0037] In a preferred embodiment of the present invention, the authentication information includes personal identification, contact information, biometrics, and delivery address;

[0038] In a preferred embodiment of the present invention, the process of acquiring high-frequency e-commerce users includes:

[0039] Obtain the total number of all transaction tasks in the historical transaction data, denoted as the total transaction volume, and obtain the number of transaction tasks participated in by cross-border e-commerce users in the historical transaction data, denoted as the user transaction volume. The participation rate of cross-border e-commerce users is Pop = n / N × 100%, where n is the user transaction volume and N is the total transaction volume.

[0040] If the participation rate of the cross-border e-commerce user is greater than or equal to a preset threshold, then the cross-border e-commerce user is a high-frequency e-commerce user, and the threshold is set within the range of [50%, 90%]; otherwise, the cross-border e-commerce user is not a high-frequency e-commerce user.

[0041] Understandably, the transaction processing center executes cross-border e-commerce user transaction tasks and acquires historical transaction data, which includes transaction tasks and associated authentication information, providing a data foundation for subsequent analysis. Based on historical transaction data, high-frequency e-commerce users are selected, and their authentication information is marked as high-frequency authentication information. The principle is that high-frequency users transact frequently, and their authentication information is used frequently. Focusing on them can optimize subsequent transaction processes.

[0042] Step S2: Establish the time series of the high-frequency authentication information, select several time nodes on the time series, obtain the text vector of the high-frequency authentication information at each time node, and obtain a set of text vectors;

[0043] Based on principal component analysis, feature vectors of the text vector set are obtained, and projection values ​​of each text vector relative to the feature vectors are obtained; based on the projection values ​​of the high-frequency authentication information at each time point, text change curves of the high-frequency authentication information are established.

[0044] Understandably, by presenting the changing trends of high-frequency authentication information, an intuitive data basis is provided for subsequent judgment on whether high-frequency authentication information is inert information, so as to determine the stable range of information, and then realize operations such as pre-caching of information within the stable range, thereby improving the efficiency and security of cross-border e-commerce transactions.

[0045] In a preferred embodiment of the present invention, the process of establishing the time series of the high-frequency authentication information includes:

[0046] The time of the first upload of the high-frequency authentication information in the historical transaction data is obtained and recorded as the first upload time, and the time of the latest upload of the high-frequency authentication information is obtained and recorded as the latest upload time; the time period between the first upload time and the latest upload time is recorded as the time series of the high-frequency authentication information.

[0047] In a preferred embodiment of the present invention, a time interval threshold is set, and several time nodes are selected at equal intervals on the time series according to the time interval threshold.

[0048] In a preferred embodiment of the present invention, the text vector of the high-frequency authentication information is obtained based on natural language processing technology;

[0049] In a preferred embodiment of the present invention, the process of obtaining the feature vectors of the text vector set includes:

[0050] Obtain a set of text vectors {(x1, y1), (x2, y2), ..., (xm, ym)}, where (xm, ym) represents the coordinates of the m-th text vector, and m is the total number of text vectors; based on the set of text vectors, obtain the mean of the set of text vectors along the x-axis. and the mean in the y-dimension Where xi represents the coordinate value of the i-th text vector in the x-dimensional direction, yi represents the coordinate value of the i-th text vector in the y-dimensional direction, i∈[1,m] and i is a positive integer;

[0051] And obtain the standard deviation of the text vector set in the x-dimensional dimension. and the standard deviation in the y-dimension

[0052] Then we obtain the standardized value of the i-th text vector in the x-dimensional direction. and standardized values ​​in the y-dimension Obtain the covariance matrix of the text vector set. The maximum eigenvalue of the text vector set is obtained based on the covariance matrix, and the eigenvector corresponding to the maximum eigenvalue is obtained.

[0053] In a preferred embodiment of the present invention, the process of obtaining the projection value of the feature vector includes:

[0054] Let the feature vector be (a, b). For any text vector (xi, yi), the projection value of the text vector (xi, yi) is z = a*xi′ + b*yi′.

[0055] In a preferred embodiment of the present invention, the process of establishing the text change curve of the high-frequency authentication information includes:

[0056] Each time node is numbered, and a rectangular coordinate system is established with the time node number as the horizontal axis and the projection value as the vertical axis; each numbered time node and its corresponding projection value are converted into coordinate points at corresponding positions in the rectangular coordinate system; each coordinate point is connected by a smooth curve, and the curve is recorded as the text change curve.

[0057] It is understandable that high-frequency authentication information is arranged in chronological order to form a time series, which makes it easier to observe how the information changes over time. Time nodes are selected on the time series, and the high-frequency authentication information is converted into text vectors. The text vector is a data form that can quantitatively represent text features in mathematical space. For example, the text information is quantified using methods such as the bag-of-words model and TF-IDF to obtain a set of text vectors.

[0058] PCA (Programmatical Conversion Analysis) is used to process the text vector set. PCA transforms multiple related variables into a few uncorrelated composite variables (principal components) through linear transformation. These principal components can retain the information of the original data to the greatest extent. The feature vectors of the text vector set are calculated, and then the projection values ​​of each text vector relative to the feature vectors are obtained. These projection values ​​reflect the changes of high-frequency authentication information in different principal component directions. Based on the projection values, a text change curve is established to intuitively show the trend of high-frequency authentication information over time.

[0059] Step S3: Based on the text change curve, obtain the volatility of the high-frequency authentication information, and based on the volatility, determine whether the high-frequency authentication information is inert change information;

[0060] If the high-frequency authentication information is inertial change information, then determine the stable range of the inertial change information; establish a pre-cache center, and store the inertial change information in the latest stable range in the pre-cache center in real time;

[0061] In a preferred embodiment of the present invention, the process of obtaining the volatility of the high-frequency authentication information includes:

[0062] Several time window values ​​are set, each time window value being a time period composed of several time nodes, with different time window values ​​containing different numbers of time nodes; based on the time window values, several time windows are extracted from the time series, and the curve segment corresponding to each time window on the text change curve is obtained;

[0063] Obtain the number of extreme points on the curve segment, excluding the two endpoints of the curve segment; based on the number of extreme points on all curve segments, obtain the average number of extreme points, which is denoted as the average number of extreme points corresponding to the time window value;

[0064] Based on the average number of extreme points corresponding to each time window value, a regression line is obtained by fitting using the least squares method, and the slope of the regression line is obtained and denoted as the volatility of the text change curve.

[0065] Understandably, different time window values ​​are set to extract time windows from the time series; different time windows contain different numbers of time nodes, aiming to observe changes in high-frequency authentication information from different time scales; short windows capture short-term fluctuations, while long windows reflect long-term trends, and multi-scale analysis provides a more comprehensive portrayal of information changes; the number of extreme points (excluding endpoints) on the curve segment corresponding to each time window is obtained; extreme points reflect the turning points of curve changes, with more extreme points indicating frequent changes in information during that period, and vice versa; by counting the number of extreme points, the degree of change of high-frequency authentication information within each time window is quantified; and all curves are calculated. The mean number of extreme points on a segment is used to obtain the mean number of extreme points corresponding to each time window value. This mean comprehensively reflects the frequency of information changes within the corresponding time window, eliminates the interference of fluctuations in individual curve segments, and makes the data more representative. Based on the least squares method, a regression line is fitted according to the mean number of extreme points corresponding to each time window value. The least squares method minimizes the sum of squared errors between the regression line and each data point, finding the best linear relationship between the mean number of extreme points and the time window. The slope of the regression line, i.e., the volatility, reflects the overall trend and degree of change of high-frequency authentication information over time. A large slope indicates rapid change, while a small slope indicates slow change.

[0066] In a preferred embodiment of the present invention, the process of determining whether the high-frequency authentication information is inertial change information includes:

[0067] A fluctuation threshold is set. If the fluctuation rate of the text change curve is less than the fluctuation threshold, then the high-frequency authentication information corresponding to the text change curve is inert change information; otherwise, the high-frequency authentication information corresponding to the text change curve is not inert change information.

[0068] Understandably, by calculating the volatility of the text change curve as described above, the changes of high-frequency authentication information over time are quantified. Volatility is a numerical indicator that reflects the frequency and magnitude of changes in high-frequency authentication information over time. For example, the greater the volatility, the more drastic and frequent the information changes; the smaller the volatility, the slower and more stable the information changes.

[0069] The volatility of the calculated text change curve is compared with a set volatility threshold. If the volatility is less than the threshold, it means that the high-frequency authentication information changes at a relatively low level over time, and the change is slow and stable, which meets the characteristics of "inertial change" and is therefore identified as inertial change information. Conversely, if the volatility is greater than or equal to the threshold, it indicates that the high-frequency authentication information changes more frequently or drastically, and does not meet the conditions for inertial change information, so it is not identified as inertial change information. This comparison method is used to classify whether high-frequency authentication information is inertial change information, providing a basis for subsequent operations such as determining the stable interval and pre-caching.

[0070] In a preferred embodiment of the present invention, the process of determining the stable interval of the inertial change information includes:

[0071] Several reference points are selected equally on the text change curve, and the slope of the tangent line of the text change curve at each reference point is obtained; the slope difference K between the reference point and the previous reference point is obtained in real time, which is K = |kw-k(w-1)|, where kw represents the k-th reference point and k(w-1) represents the (k-1)-th reference point; if there are several consecutive reference points on the text change curve where the slope difference K is 0, the horizontal coordinate interval formed by these reference points is recorded as the stable interval;

[0072] In a preferred embodiment of the present invention, the process of obtaining the pre-cache center includes:

[0073] Obtain all stable intervals on the text change curve, and obtain the length of the horizontal coordinate interval of each stable interval. Obtain the minimum value of the length of each horizontal coordinate interval, and record it as the stable duration of the inertial change information.

[0074] The text change curve containing the inertial change information is obtained in real time and denoted as the current text change curve. Several horizontal coordinate intervals on the current text change curve, excluding the stable interval, are denoted as change intervals. The latest time node is obtained and denoted as the current time node. The nearest change interval before the current time node is obtained and denoted as the most recent change interval. The end time of the most recent change interval is obtained, and the length of the time period between the end time and the current time node is obtained. The end time is the last time node on the most recent change interval. If the time length is less than the stable duration, the inertial change information at the end time is obtained and stored in the pre-cache center.

[0075] It is understandable that by calculating the slope difference between adjacent reference points, the change rate of high-frequency authentication information can be measured; if the slope difference is large, it indicates that the information change rate fluctuates greatly; if the slope difference is small, it indicates that the change rate is relatively stable.

[0076] When there are several consecutive reference points with slope differences K all of 0, it means that the rate of change of high-frequency authentication information remains constant and does not change within the time period corresponding to these reference points. From the perspective of actual business, this means that the high-frequency authentication information is in a relatively stable state within this time period. Therefore, the horizontal coordinate interval (corresponding time interval) formed by these reference points is determined as the stable interval, providing a time range basis for subsequent operations such as pre-caching of lazy change information.

[0077] Step S4: When the cross-border e-commerce user performs a transaction task through the transaction processing center, the information extraction time and the transaction time period are obtained; when the transaction task involves lazy change information, if the information extraction time and the transaction time period are in the same stable range, the transaction processing center directly obtains the lazy change information from the pre-cached center.

[0078] In a preferred embodiment of the present invention, the information extraction time is the moment when the transaction processing center extracts the authentication information of the cross-border e-commerce user when it performs a transaction task.

[0079] Set the shortest transaction duration T, and denote the information extraction time as t. Then, the transaction end time t' = t + T is obtained. The time period consisting of the information extraction time and the transaction end time is denoteed as the transaction time period.

[0080] In a preferred embodiment of the present invention, the most recent end time before the information extraction time is obtained, denoted as the most recent change time T', and the current stable interval [T', T'+T] is obtained; if the following conditions are met... If the information extraction time and the transaction time period are within the same stable range, then the information extraction time and the transaction time period are not within the same stable range. In this case, the transaction processing center prompts the cross-border e-commerce user to re-upload the inertial change information.

[0081] Understandably, the timeliness and stability of inertial change information are determined by judging whether the information extraction time and the transaction period are both within a stable range; if they are both within a stable range, it means that the information has not changed since the last information change and has remained stable during this transaction period.

[0082] If the information extraction time and the transaction time period are within the same stable range, it indicates that the inertial change information is effective and stable in this transaction. The transaction processing center can directly obtain this information from the pre-cached center to avoid repeated collection and verification and improve transaction efficiency.

[0083] If the data is not within the same stable range, it means that the inertia change information may have changed. To ensure transaction security and data accuracy, the transaction processing center will prompt users to re-upload the inertia change information to ensure that subsequent transactions use the latest and most accurate information.

[0084] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A multi-party secure transaction method based on cross-border e-commerce business rules, characterized in that, Includes the following steps: Step S1: Establish a transaction processing center, obtain historical transaction data from the transaction processing center, and filter out high-frequency e-commerce users and their high-frequency authentication information based on the historical transaction data; Step S2: Establish the time series of the high-frequency authentication information, select several time nodes in the time series, obtain the text vector of the high-frequency authentication information at each time node, and obtain the text vector set; obtain the text change curve of the high-frequency authentication information based on principal component analysis technology; Step S3: Based on the text change curve, determine whether the high-frequency authentication information is inertly changing information; if the high-frequency authentication information is inertly changing information, determine the stable range of the inertly changing information; establish a pre-cache center, and store the inertly changing information in the latest stable range in the pre-cache center in real time; Step S4: When a cross-border e-commerce user performs a transaction task through the transaction processing center, the information extraction time and transaction time period are obtained; when the transaction task involves lazy change information, if the information extraction time and the transaction time period are in the same stable range, the transaction processing center directly obtains the lazy change information from the pre-cached center. In step S2, the process of obtaining the text change curve of the high-frequency authentication information includes: Get the text vector set ,in Let represent the coordinates of the m-th text vector, where m is the total number of text vectors; based on the text vector set, obtain the mean of the text vector set along the x-axis. and the mean in the y-dimension , where xi represents the coordinate value of the i-th text vector in the x-dimensional direction, yi represents the coordinate value of the i-th text vector in the y-dimensional direction, i∈[1,m] and i is a positive integer; And obtain the standard deviation of the text vector set in the x-dimensional dimension. and the standard deviation in the y-dimension ; Then we obtain the standardized value of the i-th text vector in the x-dimensional direction. and standardized values ​​in the y-dimension ; Obtain the covariance matrix of the text vector set. The maximum eigenvalue of the text vector set is obtained based on the covariance matrix, and the eigenvector corresponding to the maximum eigenvalue is obtained. Let the feature vector be denoted as (a, b). For any text vector (xi, yi), the projection value of the text vector (xi, yi) is... Based on the projection values ​​of the high-frequency authentication information at each time point, establish the text change curve of the high-frequency authentication information; In step S3, the process of determining whether the high-frequency authentication information is inertial change information includes: Based on the text change curve, the volatility of the high-frequency authentication information is obtained. Based on the volatility, a volatility threshold is set. If the volatility of the text change curve is less than the volatility threshold, the high-frequency authentication information corresponding to the text change curve is inert change information; otherwise, the high-frequency authentication information corresponding to the text change curve is not inert change information. The process of obtaining the volatility of the high-frequency authentication information includes: Several time window values ​​are set, each time window value being a time period composed of several time nodes, with different time window values ​​containing different numbers of time nodes; based on the time window values, several time windows are extracted from the time series, and the curve segment corresponding to each time window on the text change curve is obtained; Obtain the number of extreme points on the curve segment, excluding the two endpoints of the curve segment; based on the number of extreme points on all curve segments, obtain the average number of extreme points, which is denoted as the average number of extreme points corresponding to the time window value; Based on the average number of extreme points corresponding to each time window value, a regression line is obtained by fitting using the least squares method, and the slope of the regression line is obtained and denoted as the volatility of the text change curve.

2. The multi-party secure transaction method based on cross-border e-commerce business rules according to claim 1, characterized in that, In step S1, the transaction processing center is used to execute transaction tasks between various cross-border e-commerce users. The historical transaction data includes the transaction tasks of each cross-border e-commerce user and the authentication information of the cross-border e-commerce users involved in each transaction task.

3. The multi-party secure transaction method based on cross-border e-commerce business rules according to claim 2, characterized in that, In step S1, the process of acquiring the high-frequency e-commerce user includes: Obtain the total number of all transaction tasks in the historical transaction data, denoted as the total transaction volume, and obtain the number of transaction tasks participated in by cross-border e-commerce users in the historical transaction data, denoted as the user transaction volume. The participation rate of cross-border e-commerce users is Pop = n / N × 100%, where n is the user transaction volume and N is the total transaction volume. If the participation rate of the cross-border e-commerce user is greater than or equal to a preset threshold, then the cross-border e-commerce user is a high-frequency e-commerce user, and the threshold is set within the range of [50%, 90%]; otherwise, the cross-border e-commerce user is not a high-frequency e-commerce user; and the authentication information of the high-frequency e-commerce user is recorded as high-frequency authentication information.

4. The multi-party secure transaction method based on cross-border e-commerce business rules according to claim 1, characterized in that, In step S2, the process of establishing the time series of the high-frequency authentication information includes: The time of the first upload of the high-frequency authentication information in the historical transaction data is obtained and recorded as the first upload time, and the time of the latest upload of the high-frequency authentication information is obtained and recorded as the latest upload time; the time period between the first upload time and the latest upload time is recorded as the time series of the high-frequency authentication information.

5. A multi-party secure transaction method based on cross-border e-commerce business rules according to claim 1, characterized in that, The process of determining the stable interval of the inertial change information includes: Several reference points are selected equally on the text change curve, and the slope of the tangent line of the text change curve at each reference point is obtained; the slope difference K = |kw-k(w-1)| between the reference point and the previous reference point is obtained in real time, where kw represents the k-th reference point and k(w-1) represents the (k-1)-th reference point; if there are several consecutive reference points on the text change curve where the slope difference K is 0, the horizontal coordinate interval formed by these reference points is recorded as the stable interval.

Images