Eye track classification anti-fraud method and device, electronic equipment and storage medium
By performing curvature segmentation, feature extraction and clustering of user gaze trajectories, and training classification models with the adaboost algorithm, the problem of unstable fraud risk identification effect in the existing technology is solved, and efficient and accurate anti-fraud is achieved in the classification of gaze trajectory.
Patent Information
- Application Number
- CN202411676644.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-21
- Publication Date
- 2025-02-07
AI Technical Summary
The existing technology has unstable effects in fraud risk identification, traditional methods have the risks of delay and camouflage attacks, and the real-time identification scheme based on user characteristics is not effective.
By obtaining the gaze trajectory sample, using the curvature slicing algorithm to segment it into sub-samples, extracting shapelet features, clustering and filtering, generating a classification feature set, and training the classification model through the Adaboost iterative algorithm to achieve real-time fraud risk identification.
The identification effect of fraud risk identification based on user gaze trajectory is improved, and a gaze trajectory classification model with good classification effect, high accuracy and strong interpretability is achieved.
Smart Images
Figure CN119397361A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of biometric identification and financial data classification, and in particular to an eye track classification anti-fraud method, device, electronic device and computer-readable storage medium. Background Art
[0002] Fraud risk identification is a long-standing and difficult area in the field of risk control. Traditional fraud risk identification usually uses a technical solution that analyzes related information such as text, browsing history, and device IP. This solution has a certain delay and is easily camouflaged and hacked. In the prior art, solutions for real-time fraud risk identification based on user characteristics usually use technical solutions such as micro-expression recognition or tone of voice recognition, but the effects of the above solutions are unstable. Summary of the invention
[0003] The present invention provides an eye tracking classification anti-fraud method, device, electronic device and computer-readable storage medium, the main purpose of which is to improve the recognition effect of fraud risk identification based on user eye tracking.
[0004] To achieve the above object, the present invention provides an eye track classification anti-fraud method, the method comprising:
[0005] Obtaining a gaze trajectory sample, and dividing the gaze trajectory sample into a plurality of sub-samples using a preset curvature segmentation algorithm;
[0006] Extracting features from the plurality of subsamples to obtain a plurality of groups of subsample features, clustering the plurality of groups of subsample features and calculating cluster centers, screening a preset number of subsample features based on a preset threshold and a distance from the cluster center to generate a classification feature set, and training and learning the classification feature set through an iterative algorithm to obtain a classification model;
[0007] The real-time gaze trajectory of the current user is obtained, the real-time gaze trajectory of the current user is classified based on the classification model, and whether the current user has a fraud risk is determined according to the classification result.
[0008] Optionally, dividing the gaze trajectory sample into a plurality of sub-samples by using a preset curvature segmentation algorithm includes:
[0009] Calculating the curvature of the gaze trajectory sample, and locating the curvature extreme point of the gaze trajectory sample based on the curvature;
[0010] The gaze trajectory sample is divided into a plurality of sub-samples based on the curvature extreme point.
[0011] Optionally, the sub-sample features are shape l ets features, including gradient direction features, gradient modulus features and curvature features.
[0012] Optionally, extracting features from the multiple groups of sub-samples to obtain multiple groups of sub-sample features includes:
[0013] Extract the gradient direction features, gradient modulus features and curvature features of each sub-sample;
[0014] The gradient direction feature, gradient modulus feature and curvature feature corresponding to each sub-sample are taken as a set of sub-sample features.
[0015] Optionally, clustering the multiple groups of sub-sample features and calculating the cluster center, and screening a preset number of sub-sample features based on a preset threshold and a distance from the cluster center to generate a classification feature set, includes:
[0016] Clustering the gradient direction features in the multiple groups of sub-sample features and calculating the first cluster center, arranging the multiple groups of sub-sample features in ascending order of distance from the first cluster center, and selecting the first N groups of sub-sample features as the first sub-sample feature set, where N is a preset threshold;
[0017] Clustering the gradient modulus features in the multiple groups of sub-sample features and calculating the second cluster center, arranging the multiple groups of sub-sample features in ascending order of distance from the second cluster center, selecting the first N groups of sub-sample features as the second sub-sample feature set, where N is a preset threshold;
[0018] Clustering the curvature features in the multiple groups of sub-sample features and calculating the third cluster center, arranging the multiple groups of sub-sample features in ascending order of distance from the third cluster center, selecting the first N groups of sub-sample features as the third sub-sample feature set, where N is a preset threshold;
[0019] A union of the first sub-sample feature set, the second sub-sample feature set and the third sub-sample feature set is extracted as a classification feature set.
[0020] Optionally, the training and learning the classification feature set through an iterative algorithm to obtain a classification model includes:
[0021] Constructing corresponding decision tree classifiers based on each group of sub-sample features in the classification feature set;
[0022] Based on the gaze trajectory samples, the decision tree classifiers corresponding to the features of each sub-sample are trained by the adaboost iterative algorithm to obtain the weight of each decision tree classifier;
[0023] Based on the weight of each decision tree classifier, all decision tree classifiers are weighted cascaded to obtain the classification model.
[0024] Optionally, classifying the real-time gaze trajectory of the current user based on the classification model, and judging whether the current user has a fraud risk according to the classification result, includes:
[0025] Inputting the real-time gaze trajectory of the current user into the classification model to obtain a classification result corresponding to the real-time gaze trajectory of the current user;
[0026] If the classification result is a positive sample, then the current user does not have a fraud risk;
[0027] If the classification result is a negative sample, then the current user is at risk of fraud.
[0028] In order to solve the above problems, the present invention also provides an eye track classification anti-fraud device, the device comprising:
[0029] A track segmentation module, used for obtaining a gaze track sample, and segmenting the gaze track sample into a plurality of sub-samples by a preset curvature segmentation algorithm;
[0030] A model training module is used to extract features from the plurality of sub-samples to obtain a plurality of groups of sub-sample features, cluster the plurality of groups of sub-sample features and calculate cluster centers, screen a preset number of sub-sample features based on a preset threshold and a distance from the cluster center to generate a classification feature set, and train and learn the classification feature set through an iterative algorithm to obtain a classification model;
[0031] The classification anti-fraud module is used to obtain the real-time gaze trajectory of the current user, classify the real-time gaze trajectory of the current user based on the classification model, and determine whether the current user has a fraud risk according to the classification result.
[0032] In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising:
[0033] at least one processor; and,
[0034] a memory communicatively connected to the at least one processor; wherein,
[0035] The memory stores instructions that can be executed 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 the above-mentioned gaze trajectory classification anti-fraud method.
[0036] In order to solve the above-mentioned problem, the present invention also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the above-mentioned gaze trajectory classification anti-fraud method is implemented.
[0037] The present invention realizes fraud risk identification by classifying the gaze trajectory of user behavior, obtains the interpretable features of the gaze trajectory based on Shape Lets feature extraction, screens weak classifiers with more robust features through a clustering algorithm, and combines the weak classifiers into a strong classifier through an AdaBoost iterative algorithm, thereby obtaining a Shape Let with good classification effect, high classification accuracy and strong interpretability, which is cascaded into a strong classifier as a classification model, thereby improving the recognition effect of fraud risk identification based on user gaze trajectory. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] Figure 1 This is a flow chart of an embodiment of the eye tracking classification anti-fraud method of the present invention;
[0039] Figure 2 A schematic diagram of an electronic device according to an embodiment of the present invention;
[0040] Figure 3 Schematic diagram of an embodiment of the gaze trajectory classification anti-fraud device of the present invention.
[0041] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0042] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0043] It should be noted that the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.
[0044] like Figure 1 , which is a flow chart of an embodiment of the gaze trajectory classification anti-fraud method of the present invention. The gaze trajectory classification anti-fraud method is applied to a gaze trajectory classification anti-fraud device, and includes steps S1-S3.
[0045] S1. Obtain a gaze trajectory sample, and divide the gaze trajectory sample into multiple sub-samples using a preset curvature segmentation algorithm.
[0046] Specifically, the gaze trajectory samples include positive samples and negative samples. The gaze trajectory samples are gaze trajectories of users when operating a financial system, acquired by the financial risk control platform.
[0047] Specifically, dividing the gaze trajectory sample into multiple sub-samples using a preset curvature segmentation algorithm includes: calculating the curvature of the gaze trajectory sample, locating the curvature extreme point of the gaze trajectory sample based on the curvature; and dividing the gaze trajectory sample into multiple sub-samples based on the curvature extreme point.
[0048] In one embodiment, the gaze trajectory sample is a time series trajectory, and the curvature calculation formula of the gaze trajectory sample is:
[0049]
[0050]
[0051] Among them, t is time, x and y are the coordinates of the time series trajectory with respect to time t, a1, a2, a3 and b1, b2, b3 are preset constants, and k is the curvature.
[0052] S2. Extract features from the multiple sub-samples to obtain multiple groups of sub-sample features, cluster the multiple groups of sub-sample features and calculate cluster centers, filter a preset number of sub-sample features based on a preset threshold and the distance from the cluster center to generate a classification feature set, and train and learn the classification feature set through an adaboost iterative algorithm to obtain a classification model.
[0053] In one embodiment, the sub-sample features are shapelets features, including gradient direction features, gradient modulus features, and curvature features. Specifically, extracting features from the multiple groups of sub-samples to obtain multiple groups of sub-sample features includes: extracting the gradient direction features, gradient modulus features, and curvature features of each sub-sample, and taking the gradient direction features, gradient modulus features, and curvature features corresponding to each sub-sample as a group of sub-sample features.
[0054] Specifically, clustering the multiple groups of sub-sample features and calculating the cluster centers, and screening a preset number of sub-sample features based on a preset threshold and a distance from the cluster center to generate a classification feature set includes: clustering the gradient direction features in the multiple groups of sub-sample features and calculating a first cluster center, arranging the multiple groups of sub-sample features in ascending order of distance from the first cluster center, screening the first N groups of sub-sample features as the first sub-sample feature set, where N is a preset threshold; clustering the gradient modulus length features in the multiple groups of sub-sample features and calculating a second cluster center, Arrange the multiple groups of sub-sample features according to the distance from the second cluster center in ascending order, select the first N groups of sub-sample features as the second sub-sample feature set, N is a preset threshold; cluster the curvature features in the multiple groups of sub-sample features and calculate the third cluster center, arrange the multiple groups of sub-sample features according to the distance from the third cluster center in ascending order, select the first N groups of sub-sample features as the third sub-sample feature set, N is a preset threshold; extract the union of the first sub-sample feature set, the second sub-sample feature set and the third sub-sample feature set as the classification feature set.
[0055] Specifically, the classification model is obtained by training and learning the classification feature set through the adaboost iterative algorithm, including: constructing corresponding decision tree classifiers based on each group of sub-sample features in the classification feature set; training the decision tree classifiers corresponding to each sub-sample feature through the adaboost iterative algorithm based on the gaze trajectory sample to obtain the weight of each decision tree classifier; and weighted cascading all decision tree classifiers based on the weight of each decision tree classifier to obtain the classification model.
[0056] S3. Obtain the real-time gaze trajectory of the current user, classify the real-time gaze trajectory of the current user based on the classification model, and determine whether the current user has a fraud risk according to the classification result.
[0057] Specifically, classifying the real-time gaze trajectory of the current user based on the classification model, and judging whether the current user is at risk of fraud according to the classification result includes: inputting the real-time gaze trajectory of the current user into the classification model to obtain a classification result corresponding to the real-time gaze trajectory of the current user; if the classification result is a positive sample, the current user is not at risk of fraud; if the classification result is a negative sample, the current user is at risk of fraud.
[0058] It can be seen from the above embodiments that the embodiments of the present invention realize fraud risk identification by classifying the gaze trajectory of user behavior, obtains the interpretable features of the gaze trajectory based on shape l ets feature extraction, screens weak classifiers with more robust features through a clustering algorithm, and combines the weak classifiers into a strong classifier through the adaboost iterative algorithm, thereby obtaining a shape l et cascaded into a strong classifier with good classification effect, high classification accuracy and strong interpretability as a classification model, thereby improving the recognition effect of fraud risk identification based on user gaze trajectory.
[0059] like Figure 2 , which is a schematic diagram of an embodiment of an electronic device of the present invention. The electronic device 1 is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions. The electronic device 1 can be a computer, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of hosts or network servers, wherein cloud computing is a type of distributed computing, a super virtual computer composed of a group of loosely coupled computer sets.
[0060] In this embodiment, the electronic device 1 includes, but is not limited to, a memory 11, a processor 12, and a network interface 13, which can be interconnected for communication via a system bus. The memory 11 stores a gaze trajectory classification anti-fraud program 10, and the gaze trajectory classification anti-fraud program 10 can be executed by the processor 12. Figure 3 Only the electronic device 1 having components 11-13 and the gaze tracking classification anti-fraud program 10 is shown. It can be understood by those skilled in the art that Figure 3 The structure shown does not constitute a limitation on the electronic device 1 , and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.
[0061] Among them, the memory 11 includes a memory and at least one type of readable storage medium. The memory provides a cache for the operation of the electronic device 1; the readable storage medium can be volatile or non-volatile. Specifically, the readable storage medium can be a storage medium such as a flash memory, a hard disk, a multimedia card, a card-type memory (for example, SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the readable storage medium can be an internal storage unit of the electronic device 1, such as a hard disk of the electronic device 1; in other embodiments, the storage medium can also be an external storage device of the electronic device 1, such as a plug-in hard disk equipped on the electronic device 1, a smart memory card (SmartMediaCard, SMC), a secure digital (SecureDigital, SD) card, a flash card (FlashCard), etc. In this embodiment, the readable storage medium of the memory 11 mainly includes a program storage area and a data storage area, wherein the program storage area is generally used to store an operating system and various application software installed on the electronic device 1, such as storing the code of the gaze track classification anti-fraud program 10 in an embodiment of the present invention; the data storage area can store data created according to the use of the blockchain node, such as various types of data that have been output or are to be output.
[0062] In some embodiments, the processor 12 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 12 is generally used to control the overall operation of the electronic device 1, such as performing control and processing related to data interaction or communication with other devices. In this embodiment, the processor 12 is used to run the program code stored in the memory 11 or process data, such as running the eye tracking classification anti-fraud program 10.
[0063] The network interface 13 may include a wireless network interface or a wired network interface, and the network interface 13 is used to establish a communication connection between the electronic device 1 and a client (not shown in the figure).
[0064] Optionally, the electronic device 1 may further include a user interface, which may include a display (Display), an input unit such as a keyboard (Keyboard), and the optional user interface may also include a standard wired interface and a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an organic light-emitting diode (Organic Light-Emitting Diode, OLED) touch device, etc. Among them, the display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device 1 and to display a visual user interface.
[0065] In one embodiment of the present invention, the gaze trajectory classification anti-fraud program 10 implements the following steps S1-S3 when executed by the processor 12.
[0066] S1. Obtaining a gaze trajectory sample, and dividing the gaze trajectory sample into a plurality of sub-samples using a preset curvature segmentation algorithm;
[0067] S2, extracting features from the plurality of sub-samples to obtain a plurality of groups of sub-sample features, clustering the plurality of groups of sub-sample features and calculating cluster centers, screening a preset number of sub-sample features based on a preset threshold and a distance from the cluster center to generate a classification feature set, and training and learning the classification feature set through an iterative algorithm to obtain a classification model;
[0068] S3. Obtain the real-time gaze trajectory of the current user, classify the real-time gaze trajectory of the current user based on the classification model, and determine whether the current user has a fraud risk according to the classification result.
[0069] The specific operation steps implemented in the above steps S1-S3 are substantially the same as those in the first embodiment of the gaze trajectory classification anti-fraud method of the present invention, and will not be repeated here.
[0070] In other embodiments, the gaze trajectory classification anti-fraud program 10 can also be divided into one or more modules, one or more modules are stored in the memory 11, and are executed by one or more processors (processor 12 in this embodiment) to complete the present invention. The module referred to in the present invention refers to a series of computer program instruction segments that can perform specific functions, which are used to describe the execution process of the gaze trajectory classification anti-fraud program 10 in the electronic device 1.
[0071] like Figure 3 , which is a functional module diagram of an eye tracking classification anti-fraud device provided in one embodiment of the present invention.
[0072] In the first embodiment of the present invention, the gaze trajectory classification anti-fraud device 100 includes a trajectory segmentation module 110, a model training module 120 and a classification anti-fraud module 130, exemplarily:
[0073] The track segmentation module 110 is used to obtain a gaze track sample and segment the gaze track sample into a plurality of sub-samples using a preset curvature segmentation algorithm;
[0074] The model training module 120 is used to extract features from the multiple sub-samples to obtain multiple groups of sub-sample features, cluster the multiple groups of sub-sample features and calculate cluster centers, screen a preset number of sub-sample features based on a preset threshold and a distance from the cluster center to generate a classification feature set, and train and learn the classification feature set through an iterative algorithm to obtain a classification model;
[0075] The classification anti-fraud module 130 is used to obtain the real-time gaze trajectory of the current user, classify the real-time gaze trajectory of the current user based on the classification model, and determine whether the current user has a fraud risk according to the classification result.
[0076] The specific operation steps implemented when the above-mentioned trajectory segmentation module 110, model training module 120 and classification anti-fraud module 130 are executed are substantially the same as those in the above-mentioned embodiment and will not be repeated here.
[0077] In addition, the embodiment of the present invention also proposes a computer-readable storage medium, which can be volatile or non-volatile. Specifically, the computer-readable storage medium can be any one or any combination of a hard disk, a multimedia card, an SD card, a flash memory card, an SMC, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a portable compact disk read-only memory (CD-ROM), a USB memory, etc. The computer-readable storage medium stores an eye tracking classification anti-fraud program 10.
[0078] When the gaze track classification anti-fraud program 10 is executed by the processor, the following operations are implemented:
[0079] A1. Obtain gaze trajectory samples, and divide the gaze trajectory samples into multiple sub-samples using a preset curvature segmentation algorithm;
[0080] A2. Extracting features from the plurality of subsamples to obtain a plurality of groups of subsample features, clustering the plurality of groups of subsample features and calculating cluster centers, screening a preset number of subsample features based on a preset threshold and a distance from the cluster center to generate a classification feature set, and training and learning the classification feature set through an iterative algorithm to obtain a classification model;
[0081] A3. Obtain the real-time gaze trajectory of the current user, classify the real-time gaze trajectory of the current user based on the classification model, and determine whether the current user has a fraud risk according to the classification result.
[0082] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.
[0083] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, device, article or method including the element.
[0084] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.
[0085] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A gaze trajectory classification anti-fraud method, characterized in that: The method comprises: Obtaining a gaze trajectory sample, and dividing the gaze trajectory sample into a plurality of sub-samples using a preset curvature segmentation algorithm; Extracting features from the plurality of subsamples to obtain a plurality of groups of subsample features, clustering the plurality of groups of subsample features and calculating cluster centers, screening a preset number of subsample features based on a preset threshold and a distance from the cluster center to generate a classification feature set, and training and learning the classification feature set through an iterative algorithm to obtain a classification model; The real-time gaze trajectory of the current user is obtained, the real-time gaze trajectory of the current user is classified based on the classification model, and whether the current user has a fraud risk is determined according to the classification result.
2. The gaze trajectory classification anti-fraud method according to claim 1, characterized in that: The gaze track sample is divided into a plurality of sub-samples by using a preset curvature segmentation algorithm, including: Calculating the curvature of the gaze trajectory sample, and locating the curvature extreme point of the gaze trajectory sample based on the curvature; The gaze trajectory sample is divided into a plurality of sub-samples based on the curvature extreme point.
3. The gaze trajectory classification anti-fraud method according to claim 1, characterized in that: The sub-sample features are shapelets features, including gradient direction features, gradient modulus features and curvature features.
4. The gaze trajectory classification anti-fraud method according to claim 1, characterized in that: The extracting features from the multiple groups of sub-samples to obtain multiple groups of sub-sample features includes: Extract the gradient direction features, gradient modulus features and curvature features of each sub-sample; The gradient direction feature, gradient modulus feature and curvature feature corresponding to each sub-sample are taken as a set of sub-sample features.
5. The gaze trajectory classification anti-fraud method according to claim 1, characterized in that: The step of clustering the plurality of groups of sub-sample features and calculating the cluster center, and screening a preset number of sub-sample features based on a preset threshold and a distance from the cluster center to generate a classification feature set includes: Clustering the gradient direction features in the multiple groups of sub-sample features and calculating the first cluster center, arranging the multiple groups of sub-sample features in ascending order of distance from the first cluster center, and selecting the first N groups of sub-sample features as the first sub-sample feature set, where N is a preset threshold; Clustering the gradient modulus features in the multiple groups of sub-sample features and calculating the second cluster center, arranging the multiple groups of sub-sample features in ascending order of distance from the second cluster center, selecting the first N groups of sub-sample features as the second sub-sample feature set, where N is a preset threshold; Clustering the curvature features in the multiple groups of sub-sample features and calculating the third cluster center, arranging the multiple groups of sub-sample features in ascending order of distance from the third cluster center, selecting the first N groups of sub-sample features as the third sub-sample feature set, where N is a preset threshold; A union of the first sub-sample feature set, the second sub-sample feature set and the third sub-sample feature set is extracted as a classification feature set.
6. The gaze trajectory classification anti-fraud method according to claim 1, characterized in that: The step of training and learning the classification feature set through an iterative algorithm to obtain a classification model includes: Constructing corresponding decision tree classifiers based on each group of sub-sample features in the classification feature set; Based on the gaze trajectory samples, the decision tree classifiers corresponding to the features of each sub-sample are trained by the adaboost iterative algorithm to obtain the weight of each decision tree classifier; Based on the weight of each decision tree classifier, all decision tree classifiers are weighted cascaded to obtain the classification model.
7. The gaze trajectory classification anti-fraud method according to claim 1, characterized in that: The classifying the real-time gaze trajectory of the current user based on the classification model, and judging whether the current user has a fraud risk according to the classification result, includes: Inputting the real-time gaze trajectory of the current user into the classification model to obtain a classification result corresponding to the real-time gaze trajectory of the current user; If the classification result is a positive sample, then the current user does not have a fraud risk; If the classification result is a negative sample, then the current user is at risk of fraud.
8. An eye track classification anti-fraud device, characterized in that: The device comprises: A track segmentation module, used for obtaining a gaze track sample, and segmenting the gaze track sample into a plurality of sub-samples by a preset curvature segmentation algorithm; A model training module is used to extract features from the plurality of sub-samples to obtain a plurality of groups of sub-sample features, cluster the plurality of groups of sub-sample features and calculate cluster centers, screen a preset number of sub-sample features based on a preset threshold and a distance from the cluster center to generate a classification feature set, and train and learn the classification feature set through an iterative algorithm to obtain a classification model; The classification anti-fraud module is used to obtain the real-time gaze trajectory of the current user, classify the real-time gaze trajectory of the current user based on the classification model, and determine whether the current user has a fraud risk according to the classification result.
9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; 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 the gaze track classification anti-fraud method as described in any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the gaze trajectory classification anti-fraud method as described in any one of claims 1 to 7 is implemented.