A platform transaction security detection system based on big data

Abstract

The present invention belongs to the field of security detection, and discloses a platform transaction security detection system based on big data, including a user terminal and a transaction platform; the user terminal is used to obtain conventional detection data and action detection data, and send the conventional detection data and action detection data to the transaction platform; the conventional detection data includes IP address, device MAC address and device type; the action detection data includes the click information of the user from the user opening the user terminal to the initiation of the transaction request; the transaction platform is used to judge whether the current transaction environment is safe based on the detection data and action detection data. Compared with the existing security detection projects, the present invention also obtains click information, and by performing security detection on the click information, the security of the online shopping platform can be improved.

Description

Technical Field

[0001] The present invention relates to the field of security detection, and in particular to a platform transaction security detection system based on big data. Background Art

[0002] During the transaction process on the online shopping platform, the user's account or digital certificate may be stolen. In this case, the user's account may be used by others to perform operations that are not in line with the user's wishes, such as canceling an order that has been placed, changing the delivery address of an order, etc.

[0003] The existing online shopping platform security detection generally determines whether the transaction is safe by detecting the user's commonly used IP address, commonly used device identification information and other data. However, IP addresses and commonly used device identification information can be modified, and the security is still not high enough. Summary of the invention

[0004] The purpose of the present invention is to disclose a platform transaction security detection system based on big data, so as to solve the problem that when the existing online shopping platform performs security detection based on IP addresses and identification information of commonly used devices, the security is still not high enough because both the IP addresses and the identification information of commonly used devices can be modified.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] A platform transaction security detection system based on big data, including a user end and a transaction platform;

[0007] The user end is used to obtain regular detection data and motion detection data, and send the regular detection data and motion detection data to the trading platform;

[0008] Conventional detection data includes IP address, device MAC address and device type;

[0009] Action detection data includes user click information from the time the user opens the client to the time the user initiates a transaction request;

[0010] The trading platform is used to determine whether the current trading environment is safe based on detection data and action detection data.

[0011] Optionally, the click information includes the click location and the click time.

[0012] Optionally, the user terminal includes a first collection module and a second collection module;

[0013] The first collection module is used to obtain conventional detection data;

[0014] The second collection module is used to obtain motion detection data.

[0015] Optionally, the user terminal further includes a first communication module;

[0016] The first communication module is used to send the conventional detection data and the motion detection data to the transaction platform.

[0017] Optionally, the first communication module includes a network card or a wireless radio frequency chip.

[0018] Optionally, the trading platform includes a second communication module and a storage module;

[0019] The second communication module is used to receive the conventional detection data and the motion detection data sent by the user terminal;

[0020] The storage module is used to store the conventional detection data and the motion detection data received by the second communication module.

[0021] Optionally, the storage module is also used to store the security detection model.

[0022] Optionally, the trading platform also includes a detection module;

[0023] The detection module is used to input conventional detection data and action detection data into the safety detection model for detection to obtain safety detection results.

[0024] Optionally, the trading platform further includes an input module;

[0025] The input module is used by managers with modification authority to modify the security detection rules stored in the storage module.

[0026] Optionally, the device type includes any one of mobile phones, tablet computers, laptop computers, desktop computers, and all-in-one computers.

[0027] Compared with the existing platform transaction security detection projects, the present invention also obtains click information. The clicking habits of different users vary greatly, which makes it difficult for people who illegally obtain online shopping accounts to perform illegal operations to imitate the user's original clicking habits. Therefore, by performing security detection on click information, the security of the online shopping platform can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0029] Figure 1 A schematic diagram of a platform transaction security detection system based on big data of the present invention.

[0030] Figure 2 A schematic diagram of a user terminal of the present invention. DETAILED DESCRIPTION

[0031] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0032] like Figure 1 In one embodiment shown, the present invention provides a platform transaction security detection system based on big data, including a user terminal 101 and a transaction platform 102;

[0033] The user terminal 101 is used to obtain regular detection data and motion detection data, and send the regular detection data and motion detection data to the trading platform 102;

[0034] Conventional detection data includes IP address, device MAC address and device type;

[0035] The action detection data includes click information of the user from the time when the user opens the user terminal 101 to the time when the transaction request is initiated;

[0036] The transaction platform 102 is used to determine whether the current transaction environment is safe based on the detection data and the action detection data.

[0037] Compared with the existing platform transaction security detection projects, the present invention also obtains click information. The clicking habits of different users vary greatly, which makes it difficult for people who illegally obtain online shopping accounts to perform illegal operations to imitate the user's original clicking habits. Therefore, by performing security detection on click information, the security of the online shopping platform can be improved.

[0038] Specifically, the click information includes information of a mouse click or information of a finger clicking on a touch screen.

[0039] When a user uses a device with a mouse, the click information obtained is the information of the mouse click. When a user uses a device with a touch screen, the click information obtained is the information of the finger clicking the touch screen. When a user uses a computer with touch function, the information obtained is still the mouse click information.

[0040] Specifically, the transaction request is initiated when the user adds the product to the shopping cart, clicks checkout, and prepares to pay.

[0041] Optionally, the click information includes the click location and the click time.

[0042] Specifically, when the click information is mouse click information, the click information may also include the dwell time of the mouse pointer at different positions.

[0043] Optional, such as Figure 2 As shown, the user terminal 101 includes a first collection module 111 and a second collection module 112;

[0044] The first collection module 111 is used to obtain conventional detection data;

[0045] The second collecting module 112 is used to obtain motion detection data.

[0046] Optionally, when the user uses a device with a mouse, the second collecting module 112 may include a process of executing a GetCursorPos() function to obtain the position information of the mouse through the GetCursorPos() function.

[0047] Optionally, when the user uses a device with a touch screen, the second collection module 112 may include a touch screen driver or a process that executes a getevent command.

[0048] Optionally, the user terminal 101 further includes a first communication module;

[0049] The first communication module is used to send the regular detection data and the motion detection data to the transaction platform 102 .

[0050] Optionally, the first communication module includes a network card or a wireless radio frequency chip.

[0051] Optionally, the trading platform 102 includes a second communication module and a storage module;

[0052] The second communication module is used to receive the conventional detection data and the motion detection data sent by the user terminal 101;

[0053] The storage module is used to store the conventional detection data and the motion detection data received by the second communication module.

[0054] Specifically, the trading platform 102 may include a cloud server, which includes a storage module and a second communication module. In this case, the second communication module includes a port and an API of the cloud server.

[0055] Optionally, the storage module is also used to store the security detection model.

[0056] Specifically, the security detection model is a neural network model obtained by big data training using the user's historical routine detection data and action detection data. The longer the user uses the app and the more click data is obtained, the higher the security detection accuracy of the neural network model obtained.

[0057] Optionally, the trading platform 102 further includes a detection module;

[0058] The detection module is used to input conventional detection data and action detection data into the safety detection model for detection to obtain safety detection results.

[0059] Specifically, when both the conventional detection data and the motion detection data meet the set requirements, the security detection result is that the current transaction passes the security detection. When the conventional detection data or the motion detection data does not meet the set requirements, the security detection result is that the current transaction fails the security detection.

[0060] For example, the detection items of routine detection data may include: whether the IP address is the IP address of the area where you have logged in recently, whether the MAC address of the device has changed, whether the device type has changed,

[0061] When the IP address is the IP address of the area where you have logged in frequently recently, the device MAC address has not changed, and the device type has not changed, it means that the routine detection data meets the set requirements.

[0062] To determine whether the action detection data meets the set detection requirements, the neural network model converts the click position position and click time time into a binary tuple.<position,time> The set U is compared with the comparison set cmpU calculated according to the user's historical click position and click time to obtain the correlation between the two sets. Whether the set requirements are met is determined by judging whether the correlation is greater than a set threshold.

[0063] The comparison set is a set obtained by performing linear fitting on the data in the binary set obtained according to the historical action detection data of the user.

[0064] For example, the first element in the comparison set cmpU represents the position and click time of the first click after the user opens the user terminal. The click time here is the time between opening the user terminal and the first click.

[0065] The second element can be the location and time of the second click. The click time here is the time elapsed from the first click.

[0066] Since each user has corresponding clicking habits when using the device, by judging the correlation between the latest motion detection data and the motion detection data after linear fitting, it can be determined whether they belong to the same user.

[0067] Optionally, the trading platform 102 further includes an input module;

[0068] The input module is used by managers with modification authority to modify the security detection rules stored in the storage module.

[0069] Optionally, the input module includes a biometric acquisition unit, a calculation unit, a comparison unit and an input unit;

[0070] The biometric feature acquisition unit is used to acquire the biometric feature image of the management personnel;

[0071] The computing unit is used to obtain the characteristics of the recognition area of ​​the biometric image;

[0072] The comparison unit is used to compare the features acquired by the calculation unit with the features of the identification area of ​​the biometric image of the manager with modification authority to determine whether the manager has modification authority;

[0073] The input unit is used by managers with modification authority to modify the security detection rules stored in the storage module.

[0074] Optionally, the biometric image includes a facial image or a fingerprint image.

[0075] Optionally, obtaining features of the recognition area of ​​the biometric image includes:

[0076] Performing image segmentation on the biometric image to obtain the identification area in the biometric image;

[0077] Use feature extraction algorithm to obtain features in the recognition area.

[0078] Optionally, performing image segmentation on the biometric image to obtain a recognition area in the biometric image includes:

[0079] Perform filtering calculation on the biometric image to obtain the filtered image:

[0080] Get the grayscale image imggr of the biometric image;

[0081] Get the brightness image imgli of the biometric image in Lab color space;

[0082] Perform filtering on the grayscale image imggr to obtain a first image firph;

[0083] Filter the brightness image imgli to obtain the second image sedph;

[0084] Use the following function to get the filtered image:

[0085] filph=λ×firph+(1-λ)×sedph×Φ

[0086] Among them, λ represents the filtering parameter, λ ∈ (0, 1), Φ represents the mapping coefficient, which is used to expand the value range of the pixel values in sedph from [0, 100] to [0, 255]; filph represents the filtered image;

[0087] Use a segmentation algorithm to perform image segmentation on the filtered image, obtain the region of interest in the filtered image, and use the region of interest as the recognition region.

[0088] By performing filtering calculations in different color spaces and then fusing the results of the filtering calculations, since the luminance component only contains luminance information, this enables the effective reduction of the influence of color factors on the filtering results and improves the accuracy of the filtering results.

[0089] Optionally, perform filtering processing on the grayscale image imggr to obtain the first image firph, including:

[0090] Use the NLmeans algorithm to calculate imggr to obtain firph.

[0091] Optionally, perform filtering processing on the luminance image imgli to obtain the second image sedph, including:

[0092] Use the Canny algorithm to calculate imgli to obtain an edge image;

[0093] Use the bilateral filtering algorithm to perform filtering calculations on the edge image to obtain the first filtered image;

[0094] Perform the following processing on imgli to obtain the second filtered image:

[0095] Perform wavelet decomposition on imgli to obtain wavelet high-frequency coefficients thrso and wavelet low-frequency coefficients thrst;

[0096] Use the following function to perform filtering calculations on thrso to obtain the calculated wavelet high-frequency coefficients thrsof:

[0097] If |thrso| ≥ shr, then

[0098] If |thrso| < shr, then

[0099] In the formula, w1 and w2 respectively represent the first adjustment coefficient and the second adjustment coefficient, w3 represents the filtering coefficient, w4 represents the control coefficient, shr represents the calculation threshold value, w5 represents the calculation coefficient, w6 represents the backfill coefficient, w1 ∈ (0.5, 0.6), w2 ∈ (0.1, 0.2), w3 ∈ (0.2, 0.3), w4 ∈ (2, 11), w5 ∈ (0.5, 0.6), w6 ∈ (0.2, 0.9);

[0100] Reconstruct thrsof and thrst to obtain a second filtered image;

[0101] The pixel values ​​of the pixels in the second filtered image are optimized based on the first filtered image to obtain a second image sedph.

[0102] In the process of acquiring the second image, the present invention does not directly perform filtering calculation on imgli, but acquires the first filtered image and the second filtered image respectively, and then optimizes the pixels in the second filtered image through the first filtered image, so as to improve the boundary information in the second image finally obtained. After the boundary information is improved, the accuracy of subsequent image segmentation is also improved.

[0103] Specifically, for the first filtered image, the present invention directly uses a bilateral filtering algorithm to perform filtering processing, which can effectively retain the image boundary information in the first filtered image. For the second filtered image, the present invention obtains it by performing wavelet denoising on imgli. During the wavelet denoising process, the threshold value is calculated as the wavelet high-frequency coefficient of different pixel value distributions, and the corresponding filtering function is selected for filtering calculation, thereby improving the accuracy of the filtering result.

[0104] Optionally, optimizing pixel values ​​of pixels in the second filtered image based on the first filtered image to obtain a second image sedph includes:

[0105] Obtain a set simset of pixel points in the first filtered image whose pixel values ​​are greater than a set pixel value threshold;

[0106] Obtain a set of pixel points in the second filtered image, timset, whose coordinates are the same as those of the pixel points in simset;

[0107] For the pixel point pixnd in timset, use the following function to optimize the pixel value of pixnd:

[0108]

[0109] Among them, afval pixnd Indicates the pixel value after pixnd is optimized, fval pixnd,one pixnd represents the pixel value in the second filtered image, fval pixnd,two represents the pixel value of the pixel with the same coordinate as pixnd in the first filtered image; θ( ) represents the value range control coefficient, which is used to convert afval pixnd The value is controlled in [0,255].

[0110] During the optimization process, the boundary information in the first filtered image is mainly added to the second filtered image, so that the boundary information of the second filtered image is improved.

[0111] Optionally, the device type includes any one of mobile phones, tablet computers, laptop computers, desktop computers, and all-in-one computers.

[0112] Specifically, when the device type is any one of a laptop computer, a desktop computer, and an all-in-one computer, the client is a web page or a computer application terminal of the online shopping platform.

[0113] When the device type is a mobile phone or a tablet computer, the client can be either an APP or a web page.

[0114] You can set the corresponding listening event in the web page to obtain click information.

[0115] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A platform transaction security detection system based on big data, characterized in that: Including user end and trading platform; The user end is used to obtain regular detection data and motion detection data, and send the regular detection data and motion detection data to the trading platform; Conventional detection data includes IP address, device MAC address and device type; Action detection data includes user click information from the time the user opens the client to the time the user initiates a transaction request; The trading platform is used to determine whether the current trading environment is safe based on conventional detection data and action detection data; The trading platform includes a second communication module and a storage module; The second communication module is used to receive the conventional detection data and the motion detection data sent by the user terminal; The storage module is used to store the conventional detection data and the motion detection data received by the second communication module; The storage module is also used to store security detection models; The trading platform also includes input modules; The input module is used by managers with modification authority to modify the security detection rules stored in the storage module; The input module includes a biometric acquisition unit, a calculation unit, a comparison unit and an input unit; The biometric feature acquisition unit is used to acquire the biometric feature image of the management personnel; The computing unit is used to obtain the characteristics of the identification area of ​​the biometric image; The comparison unit is used to compare the features acquired by the calculation unit with the features of the identification area of ​​the biometric image of the manager with modification authority to determine whether the manager has modification authority; The input unit is used by managers with modification authority to modify the security detection rules stored in the storage module; Obtain the characteristics of the identification area of ​​the biometric image, including: Performing image segmentation on the biometric image to obtain the identification area in the biometric image; Using feature extraction algorithm to obtain features in the identification area; Perform image segmentation on the biometric image to obtain the identification area in the biometric image, including: Perform filtering calculation on the biometric image to obtain the filtered image: Get a grayscale image of a biometric image ; Get the brightness image of the biometric image in Lab color space ; For grayscale images Perform filtering to obtain the first image ; For brightness images Perform filtering to obtain the second image ; Use the following function to get the filtered image: in, represents the filtering parameters, , Represents the mapping coefficient, which is used to The pixel value range is from Expand to ; represents the filtered image; Use a segmentation algorithm to segment the filtered image, obtain a region of interest in the filtered image, and use the region of interest as a recognition region; For brightness images Perform filtering to obtain the second image ,include: Using the Canny algorithm Perform calculations to obtain edge images; Perform filtering calculation on the edge image using a bilateral filtering algorithm to obtain a first filtered image; right Perform the following processing to obtain the second filtered image: right Perform wavelet decomposition to obtain wavelet high frequency coefficients and wavelet low frequency coefficients ; Use the following function Perform filtering calculation to obtain the calculated wavelet high-frequency coefficients : like ,but , like ,but , In the formula, and represent the first adjustment coefficient and the second adjustment coefficient respectively, represents the filter coefficient, represents the control coefficient, Indicates the calculation threshold value, represents the calculation coefficient, represents the backfill coefficient, , , , , , ; Will and Reconstructing to obtain a second filtered image; Based on the first filtered image, the pixel values ​​of the pixels in the second filtered image are optimized to obtain a second image .

2. According to a big data-based platform transaction security detection system according to claim 1, it is characterized in that: Click information includes click location and click time.

3. According to a big data-based platform transaction security detection system according to claim 1, it is characterized in that: The user end includes a first collection module and a second collection module; The first collection module is used to obtain conventional detection data; The second collection module is used to obtain motion detection data.

4. According to the big data-based platform transaction security detection system of claim 3, it is characterized in that: The user end also includes a first communication module; The first communication module is used to send the conventional detection data and the motion detection data to the transaction platform.

5. A platform transaction security detection system based on big data according to claim 4, characterized in that: The first communication module includes a network card or a wireless radio frequency chip.

6. A platform transaction security detection system based on big data according to claim 1, characterized in that: The trading platform also includes a detection module; The detection module is used to input conventional detection data and action detection data into the safety detection model for detection to obtain safety detection results.

7. The platform transaction security detection system based on big data according to claim 1 is characterized in that: Device types include mobile phones, tablet computers, laptops, desktop computers, and all-in-one computers.

Images