CN117291615A - Visual contrast analysis method and device for overcoming anti-fraud based on network payment - Google Patents
Visual contrast analysis method and device for overcoming anti-fraud based on network payment Download PDFInfo
- Publication number
- CN117291615A CN117291615A CN202311590188.5A CN202311590188A CN117291615A CN 117291615 A CN117291615 A CN 117291615A CN 202311590188 A CN202311590188 A CN 202311590188A CN 117291615 A CN117291615 A CN 117291615A
- Authority
- CN
- China
- Prior art keywords
- behavior
- payment
- fraud
- data
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000000007 visual effect Effects 0.000 title claims abstract description 191
- 238000004458 analytical method Methods 0.000 title claims abstract description 47
- 230000004927 fusion Effects 0.000 claims abstract description 75
- 238000003066 decision tree Methods 0.000 claims abstract description 59
- 238000005457 optimization Methods 0.000 claims abstract description 45
- 238000012360 testing method Methods 0.000 claims abstract description 34
- 238000010835 comparative analysis Methods 0.000 claims abstract description 4
- 238000004422 calculation algorithm Methods 0.000 claims description 38
- 239000013598 vector Substances 0.000 claims description 38
- 230000002159 abnormal effect Effects 0.000 claims description 31
- 238000009826 distribution Methods 0.000 claims description 22
- 238000000034 method Methods 0.000 claims description 21
- 238000012549 training Methods 0.000 claims description 18
- 206010000117 Abnormal behaviour Diseases 0.000 claims description 8
- 238000013139 quantization Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 claims description 5
- 238000012800 visualization Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 230000006399 behavior Effects 0.000 description 294
- 230000006870 function Effects 0.000 description 17
- 230000008569 process Effects 0.000 description 11
- 238000012545 processing Methods 0.000 description 7
- 238000013473 artificial intelligence Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000009193 crawling Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q20/00—Payment architectures, schemes or protocols
- G06Q20/38—Payment protocols; Details thereof
- G06Q20/40—Authorisation, e.g. identification of payer or payee, verification of customer or shop credentials; Review and approval of payers, e.g. check credit lines or negative lists
- G06Q20/401—Transaction verification
- G06Q20/4016—Transaction verification involving fraud or risk level assessment in transaction processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/251—Fusion techniques of input or preprocessed data
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The invention relates to the technical field of information security, and discloses a visualized comparison analysis method and a visualized comparison analysis device for overcoming anti-fraud based on network payment, wherein the visualized comparison analysis method comprises the following steps: constructing an anti-fraud decision tree model according to the historical payment data, and performing anti-fraud decision on the historical payment test data by using the anti-fraud decision tree model; generating a historical payment behavior visual view according to the payment behavior confidence level and the confidence level threshold; fusing the basic attribute data and the network payment data; extracting network payment characteristics of the multi-source fusion data, performing fraud decision on the network payment characteristics through an anti-fraud decision tree model, and generating a target behavior visual view according to the target payment behavior confidence and a confidence threshold; optimizing the visual view of the target behavior, and performing comparative analysis on network payment anti-fraud of the target user according to the visual view of the historical payment behavior and the visual view of the target optimization behavior. The invention can improve the visual contrast accuracy for overcoming anti-fraud under network payment.
Description
Technical Field
The invention relates to the technical field of information security, in particular to a visual contrast analysis method and device for overcoming anti-fraud based on network payment.
Background
Network payment is widely applied in modern life, and a convenient and quick payment mode is provided for people. Network payments, however, also face a number of fraud risks, such as fraudulent transactions, theft of identity information, hacking of accounts, etc. In order to increase the security and trustworthiness of network payment systems, a series of measures need to be taken to prevent and contain fraud risk.
Under the existing network payment, the anti-fraud technology is utilized to discover patterns and rules from massive payment data by utilizing technologies such as data mining, machine learning, data analysis and the like, and to identify fraud. However, anti-fraud techniques are not universal, and especially when fraud techniques are increasingly complex and diverse, it is not sufficient to rely on technical means alone to prevent and identify fraud, and thus the accuracy of visual comparisons against anti-fraud under network payment is low.
Disclosure of Invention
The invention provides a visual contrast analysis method and device for overcoming anti-fraud based on network payment, which mainly aim to solve the problem of lower accuracy in the process of performing visual contrast for overcoming anti-fraud based on network payment.
In order to achieve the above object, the present invention provides a visual contrast analysis method for overcoming anti-fraud based on network payment, comprising:
s1, acquiring historical payment data, constructing an anti-fraud decision tree model according to the historical payment data, and performing anti-fraud decision on historical payment test data in the historical payment data by using the anti-fraud decision tree model to obtain a payment behavior confidence coefficient;
s2, generating a historical payment behavior visible view corresponding to the historical payment data according to the payment behavior confidence level and a preset confidence level threshold;
s3, acquiring basic attribute data and network payment data of a target user, and fusing the basic attribute data and the network payment data by using a preset multi-source fusion algorithm to obtain multi-source fusion data;
s4, extracting network payment characteristics of the multi-source fusion data, performing fraud decision on the network payment characteristics through the anti-fraud decision tree model to obtain target payment behavior confidence coefficient, and generating a target behavior visual view of the target user according to the target payment behavior confidence coefficient and the confidence coefficient threshold;
s5, optimizing the target behavior visual view by using a preset anti-fraud strategy to obtain a target optimization behavior visual view, and comparing and analyzing the network payment anti-fraud of the target user according to the historical payment behavior visual view and the target optimization behavior visual view, wherein the optimizing the target behavior visual view by using the preset anti-fraud strategy to obtain the target optimization behavior visual view comprises the following steps:
S51, acquiring behavior indexes in the target behavior visual view;
s52, calculating a behavior fraud early warning value according to the behavior index by using a preset fraud early warning algorithm, wherein the fraud early warning algorithm is as follows:
wherein,is->Behavior fraud early warning value of individual behavior index, +.>As a maximum function>Is->Entropy value of individual behavior index,/->Index number, which is behavior index, < >>As a logarithmic function>Is the average value of entropy values;
s53, when the fraud early warning value is larger than a preset fraud early warning threshold value, optimizing the behavior index in the target behavior visual view by utilizing the fraud-overcoming strategy to obtain an optimized behavior index;
s54, generating the target optimization behavior visual view according to the optimization behavior index.
Optionally, the constructing an anti-fraud decision tree model according to the historical payment data includes:
extracting historical payment characteristics of the historical payment data, and generating a historical payment characteristic training set according to the historical payment characteristics;
calculating an information gain value of each payment feature node in the historical payment feature training set by using a preset information gain algorithm, wherein the information gain algorithm is as follows:
Wherein,for the information gain value,/>Sample volume in training set for historical payment features, +.>Is->Sample volume in each sample class, +.>For the number of sample categories>As a logarithmic function>Is->Sample volume in the individual historical payment feature subset, +.>Is->The +.f. in the individual historical payment feature subset training set>Sample volume in each sample class, +.>Paying for the history the number of feature sub-training sets;
selecting a payment characteristic node with the maximum information gain value as a root node, and splitting a left node and a right node on the root node;
and adding the historical payment features corresponding to the payment feature nodes into the left node and the right node one by one according to the information gain value to obtain the anti-fraud decision tree model.
Optionally, the performing anti-fraud decision on the historical payment test data in the historical payment data by using the anti-fraud decision tree model to obtain a payment behavior confidence coefficient includes:
extracting payment test characteristics of the historical payment test data;
inputting the payment test features into the anti-fraud decision tree model for feature comparison to obtain a payment test path;
And superposing the information gain value of each payment characteristic node in the payment test path to obtain the payment behavior confidence coefficient.
Optionally, the generating the historical payment behavior visual view corresponding to the historical payment data according to the payment behavior confidence and a preset confidence threshold includes:
when the confidence coefficient of the payment behavior is smaller than or equal to a preset confidence coefficient threshold value, determining an abnormal behavior distribution trend according to the historical payment characteristics of the historical payment data, and generating a fraud visible behavior view according to the abnormal behavior distribution trend;
when the confidence coefficient of the payment behavior is larger than a preset confidence coefficient threshold value, determining a normal behavior distribution trend according to the historical payment characteristics of the historical payment data, and generating a normal behavior visual view according to the normal behavior distribution trend;
and taking the fraud behavior visual view and the normal behavior visual view as historical payment behavior visual views.
Optionally, the fusing the basic attribute data and the network payment data by using a preset multi-source fusion algorithm to obtain multi-source fusion data includes:
performing numerical quantization on the basic attribute data to obtain basic attribute values, and generating basic attribute vectors according to the basic attribute values;
Performing numerical quantization on the network payment data to obtain network payment numerical values, and generating network payment vectors according to the network payment numerical values;
and fusing the basic attribute vector and the network payment vector by using the multi-source fusion algorithm to obtain multi-source fusion data, wherein the multi-source fusion algorithm is as follows:
wherein,for the multisource fusion data, +.>For the +.>Personal attribute value->For the +.>Payment attribute value->For the number of attributes of said basic attribute vector, < >>The number of payment attributes for the network payment vector.
Optionally, the extracting the network payment feature of the multisource fusion data includes:
calculating the characteristic value of the data moment of the multi-source fusion data by using the following data characteristic calculation formula:
wherein,for +.>Data moment eigenvalues of individual data, +.>The>Attribute value of individual data->For the mean value of the attribute values of all the data in the multisource fusion data,/for the mean value of the attribute values of all the data in the multisource fusion data>For the number of attributes of the basic attribute vector in the multisource fusion data, +.>For the number of payment attributes of the network payment vector in the multisource fusion data,/for the number of payment attributes of the network payment vector in the multisource fusion data >Order of central moment;
and when the data moment characteristic value is larger than a preset payment threshold value, taking the payment attribute in the multi-source fusion data vector corresponding to the data moment characteristic value as the network payment characteristic.
Optionally, the generating the target behavior visual view of the target user according to the target payment behavior confidence and the confidence threshold includes:
when the confidence coefficient of the target payment behavior is smaller than or equal to the confidence coefficient threshold value, extracting abnormal payment behavior characteristics of the target user, and generating a target fraud visual view according to the abnormal payment behavior characteristics and a preset time stamp;
when the confidence coefficient of the payment behavior is larger than the confidence coefficient threshold, extracting the normal payment behavior characteristics of the target user, and generating a target normal behavior visual view according to the normal payment behavior characteristics and a preset time stamp;
and collecting the target fraud visual image and the target normal behavior visual image as target behavior visual images of the target users.
Optionally, the comparing the network payment anti-fraud of the target user according to the historical payment behavior visual view and the target optimization behavior visual view includes:
Acquiring a target behavior numerical index in the target optimization behavior visual view;
acquiring a payment behavior numerical index in the historical payment behavior visual view;
calculating a behavior comparison index of the target behavior numerical index and the payment behavior numerical index by using the following index comparison formula:
wherein,for the behavior comparison index, < >>Is->The target behavior numerical index +_>Is->The payment behavior numerical index +_>Paying for behavior attributes in the behavior visual view;
and carrying out comparative analysis on the network payment anti-fraud of the target user according to the behavior comparison index.
Optionally, the comparing analysis of the network payment anti-fraud of the target user according to the behavior comparison index includes:
when the behavior comparison index is greater than or equal to a preset comparison index threshold, optimizing the anti-fraud strategy to obtain an optimized anti-fraud strategy;
determining an optimized behavior index by utilizing the optimization to overcome an anti-fraud strategy until the optimized behavior index is smaller than a preset comparison index threshold;
and when the behavior comparison index is smaller than a preset comparison index threshold, generating a network payment anti-fraud visualization strategy diagram of the target user.
In order to solve the above problems, the present invention further provides a visual contrast analysis device for overcoming anti-fraud based on network payment, the device comprising:
the anti-fraud decision tree model construction module is used for acquiring historical payment data, constructing an anti-fraud decision tree model according to the historical payment data, and performing anti-fraud decision on historical payment test data in the historical payment data by using the anti-fraud decision tree model to obtain a payment behavior confidence coefficient;
the historical payment behavior visual view generation module is used for generating a historical payment behavior visual view corresponding to the historical payment data according to the payment behavior confidence and a preset confidence threshold;
the data fusion module is used for acquiring basic attribute data and network payment data of a target user, and utilizing a preset multi-source fusion algorithm to fuse the basic attribute data and the network payment data to obtain multi-source fusion data;
the target behavior visual view generation module is used for extracting network payment characteristics of the multi-source fusion data, performing fraud decision on the network payment characteristics through the anti-fraud decision tree model to obtain target payment behavior confidence coefficient, and generating a target behavior visual view of the target user according to the target payment behavior confidence coefficient and the confidence coefficient threshold;
The network payment anti-fraud comparison analysis module is used for optimizing the target behavior visual view by utilizing a preset anti-fraud strategy to obtain a target optimization behavior visual view, and comparing and analyzing the network payment anti-fraud of the target user according to the historical payment behavior visual view and the target optimization behavior visual view, wherein the optimizing the target behavior visual view by utilizing the preset anti-fraud strategy to obtain the target optimization behavior visual view comprises the following steps:
acquiring behavior indexes in the target behavior visual view;
calculating a behavior fraud early warning value according to the behavior index by using a preset fraud early warning algorithm, wherein the fraud early warning algorithm is as follows:
wherein,is->Behavior fraud early warning value of individual behavior index, +.>As a maximum function>Is->Entropy value of individual behavior index,/->Index number, which is behavior index, < >>As a logarithmic function>Is the average value of entropy values;
when the fraud early warning value is larger than a preset fraud early warning threshold value, optimizing the behavior index in the target behavior visual view by utilizing the fraud-overcoming strategy to obtain an optimized behavior index;
and generating the target optimization behavior visual view according to the optimization behavior index.
According to the embodiment of the invention, the anti-fraud decision tree model is constructed through the historical payment data, and the abnormal transaction behavior and the normal transaction behavior in the historical payment data are further divided by using the anti-fraud decision tree model, so that fraud identification is facilitated for more users in network payment; the fusion of multiple data attributes of the target user is beneficial to providing more comprehensive and accurate information support for anti-fraud decisions; and carrying out fraud recognition on the payment characteristics of the target user by using the anti-fraud decision tree model, generating a behavior visual view by using the recognized payment behavior confidence, and carrying out voicing analysis on the network payment anti-fraud of the target user based on the behavior visual view, so that abnormal transaction behaviors can be more intuitively displayed. Therefore, the visualized contrast analysis method and the visualized contrast analysis device for overcoming the anti-fraud under the network payment can solve the problem of lower accuracy in the visualized contrast process of overcoming the anti-fraud under the network payment.
Drawings
FIG. 1 is a flow chart of a visual contrast analysis method for overcoming anti-fraud based on network payment according to an embodiment of the present invention;
FIG. 2 is a flow chart of a method for constructing an anti-fraud decision tree model according to an embodiment of the present invention;
FIG. 3 is a flow chart illustrating a visual view of a historical payment behavior generated according to an embodiment of the present invention;
FIG. 4 is a functional block diagram of a visual contrast analysis device for overcoming anti-fraud based on network payment according to an embodiment of the present invention;
the achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The embodiment of the application provides a visual contrast analysis method for overcoming anti-fraud based on network payment. The execution subject of the visual contrast analysis method for overcoming anti-fraud under network payment includes, but is not limited to, at least one of a server, a terminal and the like capable of being configured to execute the method provided by the embodiment of the application. In other words, the visual contrast analysis method for overcoming anti-fraud under network payment may be performed by software or hardware installed in a terminal device or a server device, and the software may be a blockchain platform. The service end includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like. The server may be an independent server, or may be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.
Referring to fig. 1, a flow chart of a visual contrast analysis method for overcoming anti-fraud based on network payment according to an embodiment of the present invention is shown. In this embodiment, the visual contrast analysis method for overcoming anti-fraud based on network payment includes:
s1, acquiring historical payment data, constructing an anti-fraud decision tree model according to the historical payment data, and performing anti-fraud decision on historical payment test data in the historical payment data by using the anti-fraud decision tree model to obtain the confidence coefficient of the payment behavior.
In the embodiment of the invention, the historical payment data comprises payment amount, payment time, payment mode, payment account, payment address, payee and the like of the target user in the network transaction process.
In detail, the historical payment data may be obtained from a pre-stored storage area including, but not limited to, a database, a blockchain, etc., by a computer sentence having a data crawling function (e.g., a Java sentence, a Python sentence, etc.).
Further, in order to prevent fraud in the online payment process, historical payment data needs to be trained to construct an anti-fraud decision tree model to master the payment transaction model and rules, and abnormal transactions are identified by the rules, so that users are prevented from being fraudulently in the online transaction process.
In the embodiment of the invention, the anti-fraud decision tree model is based on a supervised learning algorithm and is used for detecting fraud in payment transactions, and according to historical payment data, the model can automatically learn transaction modes and rules and judge whether new transactions are abnormal transactions or not based on the rules. In addition, before the anti-fraud decision tree model is constructed, data cleaning is required to be carried out on the historical payment data, invalid or repeated data is removed, missing data is filled, and the integrity and the accuracy of the data are guaranteed.
In an embodiment of the present invention, referring to fig. 2, the constructing an anti-fraud decision tree model according to the historical payment data includes:
s21, extracting historical payment characteristics of the historical payment data, and generating a historical payment characteristic training set according to the historical payment characteristics;
s22, calculating an information gain value of each payment feature node in the historical payment feature training set by using a preset information gain algorithm, wherein the information gain algorithm is as follows:
wherein,for the information gain value,/>Sample volume in training set for historical payment features, +.>Is->Sample volume in each sample class, +.>For the number of sample categories >As a logarithmic function>Is->Sample volume in the individual historical payment feature subset, +.>Is->The +.f. in the individual historical payment feature subset training set>Sample volume in each sample class, +.>Paying for the history the number of feature sub-training sets;
s23, selecting a payment characteristic node with the maximum information gain value as a root node, and splitting a left node and a right node on the root node;
and S24, adding the historical payment features corresponding to the payment feature nodes into the left node and the right node one by one according to the information gain value to obtain the anti-fraud decision tree model.
In detail, the historical payment features include payment amount, payment time, payment mode, payment account, payment address, etc., wherein the historical payment features can be extracted through feature selection, and comparison and unified processing among different payment features can be ensured through feature normalization. And constructing a historical payment feature training set according to the historical payment features corresponding to the target users.
Specifically, an information gain algorithm is utilized to calculate an information gain value of each payment feature node in a medical feature set, a payment feature node with the largest information gain value is selected as a root node, a left node and a right node are split again according to the root node, the payment feature nodes are sequentially added into the left node and the right node to obtain an anti-fraud decision tree model, if the anti-fraud decision tree model learns transaction amount and transaction places, whether a new transaction is an abnormal transaction needs to be judged, the judgment is carried out gradually based on rules of the transaction amount and the transaction places based on historical transaction modes and rules, and finally a transaction result is given.
Further, the historical payment data is divided into training data and test data, and after the anti-fraud decision tree model is trained according to the training data, the test data is required to be verified by the anti-fraud decision tree model so as to evaluate the effect of the anti-fraud decision tree model.
In the embodiment of the invention, the confidence of the payment behavior refers to the probability value of the leaf node output by the anti-fraud decision tree model, and represents the possibility that the transaction is predicted to be an abnormal transaction under a specific decision path, namely the probability value is taken as the confidence of the payment behavior.
In the embodiment of the present invention, the performing an anti-fraud decision on the historical payment test data in the historical payment data by using the anti-fraud decision tree model to obtain a payment behavior confidence coefficient includes:
extracting payment test characteristics of the historical payment test data;
inputting the payment test features into the anti-fraud decision tree model for feature comparison to obtain a payment test path;
and superposing the information gain value of each payment characteristic node in the payment test path to obtain the payment behavior confidence coefficient.
In detail, the payment test features include payment amount, payment time, payment mode, payment account, payment address and the like, wherein historical payment features can be extracted through feature selection, comparison and unified processing among different payment features are guaranteed through feature normalization, the payment test features are input into an anti-fraud decision tree model for classification, a payment test path corresponding to the payment test features is obtained, and further information gain values of each payment feature node in the payment test path are overlapped to be used as probability of leaf nodes in each payment test path, namely, payment behavior confidence degree.
Further, according to the confidence level of the payment behavior and the business requirement, corresponding processing and decision can be performed, the payment behavior is visualized, the payment mechanism and the user are helped to prevent and reduce the occurrence of fraudulent transactions, and business risks and losses are reduced.
S2, generating a historical payment behavior visible view corresponding to the historical payment data according to the payment behavior confidence level and a preset confidence level threshold.
In the embodiment of the invention, the visual view of the historical payment behavior can intuitively display the distribution and trend of the historical payment data and the detected abnormal transaction condition, such as a payment transaction histogram, the horizontal axis is time, the vertical axis is transaction amount, the distribution and change of the payment data are displayed through the histogram, and the abnormal transaction and the normal transaction can be marked according to a preset confidence threshold.
In the embodiment of the present invention, referring to fig. 3, the generating a visual view of a historical payment behavior corresponding to the historical payment data according to the confidence level of the payment behavior and a preset confidence threshold includes:
s31, when the confidence coefficient of the payment behavior is smaller than or equal to a preset confidence coefficient threshold value, determining an abnormal behavior distribution trend according to the historical payment characteristics of the historical payment data, and generating a fraud visible view according to the abnormal behavior distribution trend;
S32, when the confidence coefficient of the payment behavior is larger than a preset confidence coefficient threshold value, determining a normal behavior distribution trend according to the historical payment characteristics of the historical payment data, and generating a normal behavior visual view according to the normal behavior distribution trend;
s33, taking the fraud behavior visual view and the normal behavior visual view as historical payment behavior visual views.
In detail, when the confidence coefficient of the payment behavior is smaller than or equal to a preset confidence coefficient threshold value, the transaction is regarded as a high-risk abnormal transaction, namely, a transaction amount distribution trend corresponding to the abnormal transaction behavior is determined according to the transaction amount in the historical payment characteristics, so that an abnormal transaction behavior histogram, namely, a fraud behavior visible view, is generated according to the abnormal behavior distribution trend and the transaction time; conversely, when the confidence coefficient of the payment behavior is greater than the preset confidence coefficient threshold, the transaction is regarded as normal transaction, namely, the transaction amount distribution trend corresponding to the normal transaction behavior is determined according to the transaction amount in the historical payment feature, so that a normal transaction behavior histogram, namely, a normal behavior visual view, is generated according to the normal behavior distribution trend and the transaction time, and thus, the fraud behavior visual view and the normal behavior visual view are used as the historical payment behavior visual view corresponding to the historical payment data in the detection process through the anti-fraud decision tree model.
Further, through the visual view of the historical payment behaviors, the distribution of the historical payment data, the abnormal transaction situation and the development trend of the historical payment behaviors can be intuitively displayed, and more comprehensive and accurate information support is provided for anti-fraud decisions, so that the transaction behaviors of subsequent target users can be evaluated by using an anti-fraud decision tree model, and the transaction safety of the target users in the transaction process is ensured.
And S3, acquiring basic attribute data and network payment data of a target user, and fusing the basic attribute data and the network payment data by using a preset multi-source fusion algorithm to obtain multi-source fusion data.
In the embodiment of the invention, the basic attribute data comprise a user name, a mobile phone number, an email, a residence address, an identity card number and the like; the network payment data includes transaction amount, transaction time, transaction mode, transaction location, transaction object, etc.
In detail, the basic attribute data of the target user and the network payment data may be acquired from a pre-stored storage area including, but not limited to, a database, a blockchain, etc., through a computer sentence having a data grabbing function (e.g., java sentence, python sentence, etc.).
Further, in order to improve the quality and credibility of the data and promote the accuracy of anti-fraud decision, the basic attribute data and the network payment data can be effectively fused.
In the embodiment of the invention, the multisource fusion data is obtained by fusing a plurality of data with different sources, so that a comprehensive and accurate data set comprising basic attribute data and network payment data is obtained.
In the embodiment of the present invention, the fusing the basic attribute data and the network payment data by using a preset multi-source fusion algorithm to obtain multi-source fusion data includes:
performing numerical quantization on the basic attribute data to obtain basic attribute values, and generating basic attribute vectors according to the basic attribute values;
performing numerical quantization on the network payment data to obtain network payment numerical values, and generating network payment vectors according to the network payment numerical values;
and fusing the basic attribute vector and the network payment vector by using the multi-source fusion algorithm to obtain multi-source fusion data, wherein the multi-source fusion algorithm is as follows:
wherein,for the multisource fusion data, +.>For the +.>Personal attribute value- >For the +.>Payment attribute value->For the number of attributes of said basic attribute vector, < >>The number of payment attributes for the network payment vector.
In detail, the attribute values in the basic attribute data are represented by numerical values, and discrete data (such as gender, marital status, education degree and the like) can be quantized by adopting a Dummy Variable (Dummy Variable) coding mode, wherein 0 indicates that the data does not belong to the category, and 1 indicates that the data belongs to the category; for continuous and ordered data (such as age, income, height and the like), the continuous and ordered data can be converted into numerical data by adopting a section division mode; for non-numeric data (e.g., name, address, etc.), natural language processing techniques or string processing techniques may be employed to convert the non-numeric data into numeric data. Such as the name may be converted to an average of Unicode codes for each word. And generating a corresponding basic attribute vector by the basic attribute values, wherein if the basic attribute values corresponding to the name, the age, the sex and the ID card number in the basic attribute data are A, B, C and D, the basic attribute vector is { A, B, C and D }.
Specifically, the network payment data is also quantized to obtain a network payment value, and a network payment vector is generated according to the network payment value, for example, the network payment vector may be { E, F, G, H }, so that the basic attribute vector and the network payment vector are fused by using a multi-source fusion algorithm, i.e., the multi-source fusion data 。
Further, a more comprehensive and accurate data set can be obtained by fusing data from different sources, and the characteristics and behaviors of a target user can be better reflected, so that the accuracy and effectiveness of anti-fraud decision are improved, and the risk degree of the transaction behaviors of the target user can be detected by utilizing multi-source fusion data.
S4, extracting network payment characteristics of the multi-source fusion data, performing fraud decision on the network payment characteristics through the anti-fraud decision tree model to obtain target payment behavior confidence coefficient, and generating a target behavior visual view of the target user according to the target payment behavior confidence coefficient and the confidence coefficient threshold.
In the embodiment of the invention, the network payment characteristics refer to payment amount, payment time, payment mode, payment account, payment address and the like in the network payment data in the multi-source fusion data.
In the embodiment of the present invention, the extracting the network payment feature of the multisource fusion data includes:
calculating the characteristic value of the data moment of the multi-source fusion data by using the following data characteristic calculation formula:
wherein,for +.>Data moment eigenvalues of individual data, +. >The>Attribute value of individual data->For the mean value of the attribute values of all the data in the multisource fusion data,/for the mean value of the attribute values of all the data in the multisource fusion data>For the number of attributes of the basic attribute vector in the multisource fusion data, +.>For the number of payment attributes of the network payment vector in the multisource fusion data,/for the number of payment attributes of the network payment vector in the multisource fusion data>Order of central moment;
and when the data moment characteristic value is larger than a preset payment threshold value, taking the payment attribute in the multi-source fusion data vector corresponding to the data moment characteristic value as the network payment characteristic.
In detail, extracting network payment features from the multi-source fusion data, namely calculating a data moment feature value of each data in the multi-source fusion data to represent attributes of the payment features, calculating the degree and distribution of distance from the center in each attribute data in the multi-source fusion data one by one according to the order change of the center distance, setting the network payment features into a payment circle range according to the center point, setting the basic attribute features into a circle range according to the center point, and setting the distances between the basic attribute features and the center point into different circle ranges according to the center point, namely when the data moment feature value is larger than a preset payment threshold value, namely, the data corresponding to the data moment feature value falls into the payment circle range, and taking vector attributes in the multi-source fusion data vector corresponding to the data moment feature value as the network payment features.
Specifically, the step of performing fraud decision on the network payment feature through the fraud decision tree model to obtain the target payment behavior confidence level is consistent with the step of performing fraud decision on the historical payment test data in the historical payment data by using the fraud decision tree model in the step S1 to obtain the payment behavior confidence level, which is not described herein. The target payment behavior confidence degree is a probability value of the risk degree of the target user on the network payment behavior through an anti-fraud decision tree model according to the network payment characteristics of the target user.
Further, according to the target payment behavior confidence coefficient corresponding to the target user, a transaction behavior visual view of the target user can be generated, and further, the transaction behavior trend can be intuitively observed.
In the embodiment of the invention, the visual target behavior refers to the situation that distribution and trend of payment data of a target user and abnormal transactions are detected can be intuitively displayed.
In the embodiment of the present invention, the generating the target behavior visual view of the target user according to the target payment behavior confidence and the confidence threshold includes:
when the confidence coefficient of the target payment behavior is smaller than or equal to the confidence coefficient threshold value, extracting abnormal payment behavior characteristics of the target user, and generating a target fraud visual view according to the abnormal payment behavior characteristics and a preset time stamp;
When the confidence coefficient of the payment behavior is larger than the confidence coefficient threshold, extracting the normal payment behavior characteristics of the target user, and generating a target normal behavior visual view according to the normal payment behavior characteristics and a preset time stamp;
and collecting the target fraud visual image and the target normal behavior visual image as target behavior visual images of the target users.
In detail, when the confidence coefficient of the target payment behavior is smaller than or equal to the confidence coefficient threshold value, the transaction is regarded as a high-risk abnormal transaction, namely, an abnormal transaction behavior histogram is generated according to the transaction amount and the transaction time in the abnormal payment behavior characteristics, namely, a target fraud behavior visible view; conversely, when the target payment behavior confidence is greater than the confidence threshold, the transaction is considered as a normal transaction, that is, a normal transaction behavior histogram is generated according to the transaction amount and the transaction time in the abnormal payment behavior feature, that is, a target normal behavior visual view, so that the target fraud behavior visual view and the target normal behavior visual view are taken as target behavior visual views.
Further, through the visual view of the target behavior, the distribution of the payment data of the target user, the abnormal transaction condition and the development trend of the payment behavior can be intuitively displayed, so that corresponding anti-fraud measures can be implemented through the visual view of the behavior of the target user, and the safety of network payment is ensured.
And S5, optimizing the target behavior visual view by utilizing a preset anti-fraud strategy to obtain a target optimization behavior visual view, and comparing and analyzing the network payment anti-fraud of the target user according to the historical payment behavior visual view and the target optimization behavior visual view.
In the embodiment of the invention, the anti-fraud strategy comprises locking the account number with abnormal payment or sending out a notice or alarm in time. When the visual view of the target behavior of the target user is abnormal, timely processing the abnormal behavior by overcoming an anti-fraud strategy, and determining behavior data of the processed visual view of the target behavior, so that the visual view of the target optimized behavior is obtained.
In the embodiment of the present invention, the optimizing the target behavior visual view by using a preset anti-fraud strategy to obtain a target optimized behavior visual view includes:
acquiring behavior indexes in the target behavior visual view;
calculating a behavior fraud early warning value according to the behavior index by using a preset fraud early warning algorithm, wherein the fraud early warning algorithm is as follows:
wherein,is->Behavior fraud early warning value of individual behavior index, +. >As a maximum function>Is->Entropy value of individual behavior index,/->Index number, which is behavior index, < >>As a logarithmic function>Is the average value of entropy values;
when the fraud early warning value is larger than a preset fraud early warning threshold value, optimizing the behavior index in the target behavior visual view by utilizing the fraud-overcoming strategy to obtain an optimized behavior index;
and generating the target optimization behavior visual view according to the optimization behavior index.
In detail, the behavior indexes refer to payment amount, payment account, payment place and the like of the target user, the behavior indexes in the visual view of the target behavior can be obtained through a data crawler, and a fraud early warning value of each behavior index is calculated by using a fraud early warning algorithm, wherein the maximum value corresponding to each behavior index is selected in the fraud early warning algorithm as the fraud early warning valueEntropy of behavior indexThe probability of occurrence of the behavior indexes is determined as the probability of occurrence of the entropy of the behavior indexes by the probability of occurrence of various behavior indexes, such as the probability of occurrence of 50,500,5000 of payment amount, the probability of national or foreign of payment account number, and the probability of national or foreign of payment place.
Specifically, when the fraud early warning value is greater than the fraud early warning threshold, it indicates that the network payment behavior corresponding to the time point is abnormal, if the payment amount of network payment at 10 points is 50 yuan and the payment amount of network payment at 11 points is 50 ten thousand yuan, the payment account is locked at 11 points, and the payment amount corresponding to 11 points is set to 0, the transaction is prevented, and then an optimized behavior index is obtained, and a target optimized behavior visual view is regenerated according to the optimized behavior index.
Further, the historical payment behavior visual view and the target optimization behavior visual view are subjected to joint comparison analysis, so that payment changes in the network payment process can be mastered in time.
In the embodiment of the present invention, the comparing and analyzing the network payment anti-fraud of the target user according to the historical payment behavior visual view and the target optimization behavior visual view includes:
acquiring a target behavior numerical index in the target optimization behavior visual view;
acquiring a payment behavior numerical index in the historical payment behavior visual view;
calculating a behavior comparison index of the target behavior numerical index and the payment behavior numerical index by using the following index comparison formula:
wherein,for the behavior comparison index, < >>Is->The target behavior numerical index +_>Is->The payment behavior numerical index +_>Paying for behavior attributes in the behavior visual view;
and carrying out comparative analysis on the network payment anti-fraud of the target user according to the behavior comparison index.
In detail, comparing the target behavior numerical index in the target optimization behavior visual view with the payment behavior data index in the historical payment behavior visual view one by one, namely calculating the behavior comparison index of the target behavior numerical index and the payment behavior data index in the historical payment behavior visual view through an index comparison formula, wherein the behavior comparison index is 0 when the behavior attribute in the target optimization behavior visual view exists in the historical payment behavior visual view, the behavior values corresponding to the same behavior attribute are compared, and the behavior comparison index is obtained when the behavior attribute in the target optimization behavior visual view does not exist in the historical payment behavior visual view.
Specifically, according to the behavior comparison index, the behavior processing effect of anti-fraud on the target user by using the customer service anti-fraud strategy when the target user experiences payment fraud can be determined, and therefore, the network payment anti-fraud is subjected to comparison analysis.
In the embodiment of the present invention, the comparing and analyzing the network payment anti-fraud of the target user according to the behavior comparison index includes:
when the behavior comparison index is greater than or equal to a preset comparison index threshold, optimizing the anti-fraud strategy to obtain an optimized anti-fraud strategy;
determining an optimized behavior index by utilizing the optimization to overcome an anti-fraud strategy until the optimized behavior index is smaller than a preset comparison index threshold;
and when the behavior comparison index is smaller than a preset comparison index threshold, generating a network payment anti-fraud visualization strategy diagram of the target user.
In detail, when the behavior comparison index is greater than or equal to a preset comparison index threshold, the customer service anti-fraud strategy is required to be optimized, a notification strategy can be added, and if the customer service anti-fraud strategy is notified through sending a short message, the customer can be optimized to be a call to notify the customer, so that an optimized anti-fraud strategy is obtained, and an optimized comparison value of optimizing behavior indexes in the optimized anti-fraud strategy is recalculated until the optimized comparison value is smaller than the preset comparison index threshold; and when the behavior comparison index is smaller than a preset comparison index threshold, generating a network payment anti-fraud visual policy map of the target user, namely, after the behavior index is optimized by utilizing an optimization anti-fraud strategy, generating the anti-fraud visual policy map.
Further, through the visual contrast analysis for overcoming anti-fraud under the network payment, the user can more intuitively know the data change and trend, can find potential abnormality or abnormal change, can help the user to quickly find the abnormal situation, and performs analysis and optimization. This helps to quickly react and avoid potential risks, and through visualization tools, users can analyze the problem from different dimensions and angles, exploring the correlation and impact between different factors, thereby more fully understanding the entire anti-fraud process.
According to the embodiment of the invention, the anti-fraud decision tree model is constructed through the historical payment data, and the abnormal transaction behavior and the normal transaction behavior in the historical payment data are further divided by using the anti-fraud decision tree model, so that fraud identification is facilitated for more users in network payment; the fusion of multiple data attributes of the target user is beneficial to providing more comprehensive and accurate information support for anti-fraud decisions; and carrying out fraud recognition on the payment characteristics of the target user by using the anti-fraud decision tree model, generating a behavior visual view by using the recognized payment behavior confidence, and carrying out voicing analysis on the network payment anti-fraud of the target user based on the behavior visual view, so that abnormal transaction behaviors can be more intuitively displayed. Therefore, the visualized contrast analysis method and the visualized contrast analysis device for overcoming the anti-fraud under the network payment can solve the problem of lower accuracy in the visualized contrast process of overcoming the anti-fraud under the network payment.
Fig. 4 is a functional block diagram of a visual contrast analysis device for overcoming anti-fraud based on network payment according to an embodiment of the present invention.
The visual contrast analysis device 100 for overcoming anti-fraud based on network payment according to the present invention may be installed in an electronic device. Depending on the implementation function, the visual contrast analysis device 100 for overcoming anti-fraud under network payment may include an anti-fraud decision tree model building module 101, a historical payment behavior visual generating module 102, a data fusion module 103, a target behavior visual generating module 104 and a network payment anti-fraud contrast analysis module 105. The module of the invention, which may also be referred to as a unit, refers to a series of computer program segments, which are stored in the memory of the electronic device, capable of being executed by the processor of the electronic device and of performing a fixed function.
In the present embodiment, the functions concerning the respective modules/units are as follows:
the anti-fraud decision tree model construction module 101 is configured to obtain historical payment data, construct an anti-fraud decision tree model according to the historical payment data, and perform anti-fraud decision on historical payment test data in the historical payment data by using the anti-fraud decision tree model to obtain a payment behavior confidence coefficient;
The historical payment behavior visual generation module 102 is configured to generate a historical payment behavior visual corresponding to the historical payment data according to the payment behavior confidence level and a preset confidence level threshold;
the data fusion module 103 is configured to obtain basic attribute data and network payment data of a target user, and fuse the basic attribute data and the network payment data by using a preset multi-source fusion algorithm to obtain multi-source fusion data;
the target behavior visual view generation module 104 is configured to extract network payment features of the multi-source fusion data, perform fraud decision on the network payment features through the anti-fraud decision tree model, obtain a target payment behavior confidence level, and generate a target behavior visual view of the target user according to the target payment behavior confidence level and the confidence level threshold;
the network payment anti-fraud comparison analysis module 105 is configured to optimize the target behavior visual view by using a preset anti-fraud policy to obtain a target optimization behavior visual view, and compare and analyze the network payment anti-fraud of the target user according to the historical payment behavior visual view and the target optimization behavior visual view, where the optimizing the target behavior visual view by using the preset anti-fraud policy to obtain the target optimization behavior visual view includes:
Acquiring behavior indexes in the target behavior visual view;
calculating a behavior fraud early warning value according to the behavior index by using a preset fraud early warning algorithm, wherein the fraud early warning algorithm is as follows:
wherein,is->Behavior fraud early warning value of individual behavior index, +.>As a maximum function>Is->Entropy value of individual behavior index,/->Index number, which is behavior index, < >>As a logarithmic function>Is the average value of entropy values; />
When the fraud early warning value is larger than a preset fraud early warning threshold value, optimizing the behavior index in the target behavior visual view by utilizing the fraud-overcoming strategy to obtain an optimized behavior index;
and generating the target optimization behavior visual view according to the optimization behavior index.
In detail, each module in the visual contrast analysis device 100 for overcoming anti-fraud based on network payment in the embodiment of the present invention adopts the same technical means as the visual contrast analysis method for overcoming anti-fraud based on network payment described in fig. 1 to 3, and can produce the same technical effects, which are not described herein.
In the several embodiments provided by the present invention, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is merely a logical function division, and there may be other manners of division when actually implemented.
The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, may be located in one place, or may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional module in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units can be realized in a form of hardware or a form of hardware and a form of software functional modules.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof.
The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.
The embodiment of the application can acquire and process the related data based on the artificial intelligence technology. Among these, artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use knowledge to obtain optimal results.
Furthermore, it is evident that the word "comprising" does not exclude other elements or steps, and that the singular does not exclude a plurality. Multiple units or means as set forth in the system embodiments may also be implemented by one unit or means in software or hardware. The terms first, second, etc. are used to denote a name, but not any particular order.
Finally, it should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention.
Claims (10)
1. A visual contrast analysis method for overcoming anti-fraud based on network payment, the method comprising:
S1, acquiring historical payment data, constructing an anti-fraud decision tree model according to the historical payment data, and performing anti-fraud decision on historical payment test data in the historical payment data by using the anti-fraud decision tree model to obtain a payment behavior confidence coefficient;
s2, generating a historical payment behavior visible view corresponding to the historical payment data according to the payment behavior confidence level and a preset confidence level threshold;
s3, acquiring basic attribute data and network payment data of a target user, and fusing the basic attribute data and the network payment data by using a preset multi-source fusion algorithm to obtain multi-source fusion data;
s4, extracting network payment characteristics of the multi-source fusion data, performing fraud decision on the network payment characteristics through the anti-fraud decision tree model to obtain target payment behavior confidence coefficient, and generating a target behavior visual view of the target user according to the target payment behavior confidence coefficient and the confidence coefficient threshold;
s5, optimizing the target behavior visual view by using a preset anti-fraud strategy to obtain a target optimization behavior visual view, and comparing and analyzing the network payment anti-fraud of the target user according to the historical payment behavior visual view and the target optimization behavior visual view, wherein the optimizing the target behavior visual view by using the preset anti-fraud strategy to obtain the target optimization behavior visual view comprises the following steps:
S51, acquiring behavior indexes in the target behavior visual view;
s52, calculating a behavior fraud early warning value according to the behavior index by using a preset fraud early warning algorithm, wherein the fraud early warning algorithm is as follows:
wherein,is->Behavior fraud early warning value of individual behavior index, +.>As a maximum function>Is->Entropy value of individual behavior index,/->Index number, which is behavior index, < >>As a logarithmic function>Is the average value of entropy values;
s53, when the fraud early warning value is larger than a preset fraud early warning threshold value, optimizing the behavior index in the target behavior visual view by utilizing the fraud-overcoming strategy to obtain an optimized behavior index;
s54, generating the target optimization behavior visual view according to the optimization behavior index.
2. The visual contrast analysis method for overcoming fraud under network payment according to claim 1, wherein the constructing an anti-fraud decision tree model from the historical payment data includes:
extracting historical payment characteristics of the historical payment data, and generating a historical payment characteristic training set according to the historical payment characteristics;
calculating an information gain value of each payment feature node in the historical payment feature training set by using a preset information gain algorithm, wherein the information gain algorithm is as follows:
Wherein,for the information gain value,/>Sample volume in training set for historical payment features, +.>Is->Sample volume in each sample class, +.>For the number of sample categories>As a logarithmic function>Is->Sample volume in the individual historical payment feature subset, +.>Is->The +.f. in the individual historical payment feature subset training set>Sample volume in each sample class, +.>Paying for the history the number of feature sub-training sets;
selecting a payment characteristic node with the maximum information gain value as a root node, and splitting a left node and a right node on the root node;
and adding the historical payment features corresponding to the payment feature nodes into the left node and the right node one by one according to the information gain value to obtain the anti-fraud decision tree model.
3. The visual contrast analysis method for overcoming fraud under network payment according to claim 1, wherein the performing fraud-preventing decision on the historical payment test data in the historical payment data by using the fraud-preventing decision tree model to obtain a payment behavior confidence level comprises:
extracting payment test characteristics of the historical payment test data;
inputting the payment test features into the anti-fraud decision tree model for feature comparison to obtain a payment test path;
And superposing the information gain value of each payment characteristic node in the payment test path to obtain the payment behavior confidence coefficient.
4. The visual contrast analysis method for overcoming fraud under network payment according to claim 1, wherein the generating the historical payment behavior visual view corresponding to the historical payment data according to the payment behavior confidence and a preset confidence threshold comprises:
when the confidence coefficient of the payment behavior is smaller than or equal to a preset confidence coefficient threshold value, determining an abnormal behavior distribution trend according to the historical payment characteristics of the historical payment data, and generating a fraud visible behavior view according to the abnormal behavior distribution trend;
when the confidence coefficient of the payment behavior is larger than a preset confidence coefficient threshold value, determining a normal behavior distribution trend according to the historical payment characteristics of the historical payment data, and generating a normal behavior visual view according to the normal behavior distribution trend;
and taking the fraud behavior visual view and the normal behavior visual view as historical payment behavior visual views.
5. The visual contrast analysis method for overcoming fraud under network payment according to claim 1, wherein the fusing the basic attribute data and the network payment data by using a preset multi-source fusion algorithm to obtain multi-source fusion data comprises:
Performing numerical quantization on the basic attribute data to obtain basic attribute values, and generating basic attribute vectors according to the basic attribute values;
performing numerical quantization on the network payment data to obtain network payment numerical values, and generating network payment vectors according to the network payment numerical values;
and fusing the basic attribute vector and the network payment vector by using the multi-source fusion algorithm to obtain multi-source fusion data, wherein the multi-source fusion algorithm is as follows:
wherein,for the multisource fusion data, +.>For the +.>Personal attribute value->For the +.>Payment attribute value->For the number of attributes of said basic attribute vector, < >>The number of payment attributes for the network payment vector.
6. The visual contrast analysis method for overcoming fraud under network payment according to claim 1, wherein the extracting the network payment feature of the multisource fusion data comprises:
calculating the characteristic value of the data moment of the multi-source fusion data by using the following data characteristic calculation formula:
wherein,for +.>Data moment eigenvalues of individual data, +. >The>Attribute value of individual data->To be included in the multi-source fusion dataMean value of attribute values with data, +.>For the number of attributes of the basic attribute vector in the multisource fusion data, +.>For the number of payment attributes of the network payment vector in the multisource fusion data,/for the number of payment attributes of the network payment vector in the multisource fusion data>Order of central moment;
and when the data moment characteristic value is larger than a preset payment threshold value, taking the payment attribute in the multi-source fusion data vector corresponding to the data moment characteristic value as the network payment characteristic.
7. The visual contrast analysis method for overcoming fraud under network payment according to claim 1, wherein the generating the target behavior visual view of the target user according to the target payment behavior confidence and the confidence threshold comprises:
when the confidence coefficient of the target payment behavior is smaller than or equal to the confidence coefficient threshold value, extracting abnormal payment behavior characteristics of the target user, and generating a target fraud visual view according to the abnormal payment behavior characteristics and a preset time stamp;
when the confidence coefficient of the payment behavior is larger than the confidence coefficient threshold, extracting the normal payment behavior characteristics of the target user, and generating a target normal behavior visual view according to the normal payment behavior characteristics and a preset time stamp;
And collecting the target fraud visual image and the target normal behavior visual image as target behavior visual images of the target users.
8. The visual contrast analysis method for overcoming fraud under network payment according to claim 1, wherein the comparing the network payment fraud of the target user according to the historical payment behavior visual view and the target optimization behavior visual view comprises:
acquiring a target behavior numerical index in the target optimization behavior visual view;
acquiring a payment behavior numerical index in the historical payment behavior visual view;
calculating a behavior comparison index of the target behavior numerical index and the payment behavior numerical index by using the following index comparison formula:
wherein,for the behavior comparison index, < >>Is->The target behavior numerical index +_>Is->The payment behavior numerical index +_>Paying for behavior attributes in the behavior visual view;
and carrying out comparative analysis on the network payment anti-fraud of the target user according to the behavior comparison index.
9. The visual contrast analysis method for overcoming fraud under network payment according to claim 8, wherein the performing contrast analysis on network payment fraud of the target user according to the behavior contrast index includes:
When the behavior comparison index is greater than or equal to a preset comparison index threshold, optimizing the anti-fraud strategy to obtain an optimized anti-fraud strategy;
determining an optimized behavior index by utilizing the optimization to overcome an anti-fraud strategy until the optimized behavior index is smaller than a preset comparison index threshold;
and when the behavior comparison index is smaller than a preset comparison index threshold, generating a network payment anti-fraud visualization strategy diagram of the target user.
10. A visual contrast analysis device for overcoming anti-fraud based on network payment, the device comprising:
the anti-fraud decision tree model construction module is used for acquiring historical payment data, constructing an anti-fraud decision tree model according to the historical payment data, and performing anti-fraud decision on historical payment test data in the historical payment data by using the anti-fraud decision tree model to obtain a payment behavior confidence coefficient;
the historical payment behavior visual view generation module is used for generating a historical payment behavior visual view corresponding to the historical payment data according to the payment behavior confidence and a preset confidence threshold;
the data fusion module is used for acquiring basic attribute data and network payment data of a target user, and utilizing a preset multi-source fusion algorithm to fuse the basic attribute data and the network payment data to obtain multi-source fusion data;
The target behavior visual view generation module is used for extracting network payment characteristics of the multi-source fusion data, performing fraud decision on the network payment characteristics through the anti-fraud decision tree model to obtain target payment behavior confidence coefficient, and generating a target behavior visual view of the target user according to the target payment behavior confidence coefficient and the confidence coefficient threshold;
the network payment anti-fraud comparison analysis module is used for optimizing the target behavior visual view by utilizing a preset anti-fraud strategy to obtain a target optimization behavior visual view, and comparing and analyzing the network payment anti-fraud of the target user according to the historical payment behavior visual view and the target optimization behavior visual view, wherein the optimizing the target behavior visual view by utilizing the preset anti-fraud strategy to obtain the target optimization behavior visual view comprises the following steps:
acquiring behavior indexes in the target behavior visual view;
calculating a behavior fraud early warning value according to the behavior index by using a preset fraud early warning algorithm, wherein the fraud early warning algorithm is as follows:
wherein,is->Behavior fraud early warning value of individual behavior index, +.>As a maximum function >Is->Entropy value of individual behavior index,/->Index number, which is behavior index, < >>As a logarithmic function>Is the average value of entropy values;
when the fraud early warning value is larger than a preset fraud early warning threshold value, optimizing the behavior index in the target behavior visual view by utilizing the fraud-overcoming strategy to obtain an optimized behavior index;
and generating the target optimization behavior visual view according to the optimization behavior index.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311590188.5A CN117291615B (en) | 2023-11-27 | 2023-11-27 | Visual contrast analysis method and device for overcoming anti-fraud based on network payment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311590188.5A CN117291615B (en) | 2023-11-27 | 2023-11-27 | Visual contrast analysis method and device for overcoming anti-fraud based on network payment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117291615A true CN117291615A (en) | 2023-12-26 |
| CN117291615B CN117291615B (en) | 2024-02-06 |
Family
ID=89253920
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311590188.5A Active CN117291615B (en) | 2023-11-27 | 2023-11-27 | Visual contrast analysis method and device for overcoming anti-fraud based on network payment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117291615B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106663224A (en) * | 2014-06-30 | 2017-05-10 | 亚马逊科技公司 | Interactive interfaces for machine learning model evaluations |
| CN107680602A (en) * | 2017-08-24 | 2018-02-09 | 平安科技（深圳）有限公司 | Voice fraud recognition methods, device, terminal device and storage medium |
| CN107767138A (en) * | 2017-10-20 | 2018-03-06 | 杭州呯嘭智能技术有限公司 | On-line payment is counter to cheat method and system |
| CN110073301A (en) * | 2017-08-02 | 2019-07-30 | 强力物联网投资组合2016有限公司 | The detection method and system under data collection environment in industrial Internet of Things with large data sets |
| CN110866832A (en) * | 2019-09-30 | 2020-03-06 | 北京健康之家科技有限公司 | Risk control method, system, storage medium and computing device |
| CN111784360A (en) * | 2020-09-07 | 2020-10-16 | 北京江融信科技有限公司 | Anti-fraud prediction method and system based on network link backtracking |
| CN113744054A (en) * | 2021-09-02 | 2021-12-03 | 上海浦东发展银行股份有限公司 | Anti-fraud method, device and equipment |
| CN115809930A (en) * | 2022-12-12 | 2023-03-17 | 民太安保险公估有限公司 | Anti-fraud analysis method, device, equipment and medium based on data fusion matching |
| CN116245630A (en) * | 2022-12-29 | 2023-06-09 | 中国农业银行股份有限公司 | Anti-fraud detection method and device, electronic equipment and medium |
-
2023
- 2023-11-27 CN CN202311590188.5A patent/CN117291615B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106663224A (en) * | 2014-06-30 | 2017-05-10 | 亚马逊科技公司 | Interactive interfaces for machine learning model evaluations |
| CN110073301A (en) * | 2017-08-02 | 2019-07-30 | 强力物联网投资组合2016有限公司 | The detection method and system under data collection environment in industrial Internet of Things with large data sets |
| CN107680602A (en) * | 2017-08-24 | 2018-02-09 | 平安科技（深圳）有限公司 | Voice fraud recognition methods, device, terminal device and storage medium |
| CN107767138A (en) * | 2017-10-20 | 2018-03-06 | 杭州呯嘭智能技术有限公司 | On-line payment is counter to cheat method and system |
| CN110866832A (en) * | 2019-09-30 | 2020-03-06 | 北京健康之家科技有限公司 | Risk control method, system, storage medium and computing device |
| CN111784360A (en) * | 2020-09-07 | 2020-10-16 | 北京江融信科技有限公司 | Anti-fraud prediction method and system based on network link backtracking |
| CN113744054A (en) * | 2021-09-02 | 2021-12-03 | 上海浦东发展银行股份有限公司 | Anti-fraud method, device and equipment |
| CN115809930A (en) * | 2022-12-12 | 2023-03-17 | 民太安保险公估有限公司 | Anti-fraud analysis method, device, equipment and medium based on data fusion matching |
| CN116245630A (en) * | 2022-12-29 | 2023-06-09 | 中国农业银行股份有限公司 | Anti-fraud detection method and device, electronic equipment and medium |
Non-Patent Citations (3)
| Title |
|---|
| 杨颖;赵守香;: "互联网环境下信用卡反欺诈系统研究", 电脑知识与技术, no. 01, pages 50 - 52 * |
| 王冬一;华迎;朱峻萱;: "基于大数据技术的个人信用动态评价指标体系研究――基于社会资本视角", 国际商务(对外经济贸易大学学报), no. 01, pages 115 - 126 * |
| 王成;王昌琪;: "一种面向网络支付反欺诈的自动化特征工程方法", 计算机学报, no. 10, pages 1983 - 2001 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117291615B (en) | 2024-02-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112084383B (en) | Knowledge graph-based information recommendation method, device, equipment and storage medium | |
| CN109685647B (en) | Credit fraud detection method and training method and device of model thereof, and server | |
| CN111612038B (en) | Abnormal user detection method and device, storage medium and electronic equipment | |
| CN110163242B (en) | Risk identification method and device and server | |
| CN111612041A (en) | Abnormal user identification method and device, storage medium and electronic equipment | |
| CN113159482A (en) | Method and system for evaluating information security risk | |
| CN116305168B (en) | Multi-dimensional information security risk assessment method, system and storage medium | |
| CN113011889B (en) | Account anomaly identification method, system, device, equipment and medium | |
| CN112837069A (en) | Block chain and big data based secure payment method and cloud platform system | |
| CN111177367A (en) | Case classification method, classification model training method and related products | |
| CN114186626A (en) | Abnormity detection method and device, electronic equipment and computer readable medium | |
| CN116996325B (en) | Network security detection method and system based on cloud computing | |
| CN117235608B (en) | Risk detection method, risk detection device, electronic equipment and storage medium | |
| CN111245815B (en) | Data processing method and device, storage medium and electronic equipment | |
| CN117291615B (en) | Visual contrast analysis method and device for overcoming anti-fraud based on network payment | |
| CN115659351B (en) | Information security analysis method, system and equipment based on big data office | |
| CN115603955B (en) | Abnormal access object identification method, device, equipment and medium | |
| CN116739605A (en) | Transaction data detection method, device, equipment and storage medium | |
| CN116485185A (en) | Enterprise risk analysis system and method based on comparison data | |
| CN114861163A (en) | Abnormal account identification method, device, equipment, storage medium and program product | |
| CN115619245A (en) | Portrait construction and classification method and system based on data dimension reduction method | |
| CN115618297A (en) | Method and device for identifying abnormal enterprise | |
| CN112950222A (en) | Resource processing abnormity detection method and device, electronic equipment and storage medium | |
| CN110570301A (en) | Risk identification method, device, equipment and medium | |
| CN113822490B (en) | Asset collection method and device based on artificial intelligence and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |