CN108846737A - A kind of fraud measure and system - Google Patents
A kind of fraud measure and system Download PDFInfo
- Publication number
- CN108846737A CN108846737A CN201810353021.XA CN201810353021A CN108846737A CN 108846737 A CN108846737 A CN 108846737A CN 201810353021 A CN201810353021 A CN 201810353021A CN 108846737 A CN108846737 A CN 108846737A
- Authority
- CN
- China
- Prior art keywords
- fraud
- data
- metric
- module
- measurement
- 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
- 238000000034 method Methods 0.000 claims abstract description 68
- 238000005259 measurement Methods 0.000 claims abstract description 43
- 230000003993 interaction Effects 0.000 claims abstract description 22
- 238000012545 processing Methods 0.000 claims abstract description 8
- 238000000691 measurement method Methods 0.000 claims description 14
- 238000012795 verification Methods 0.000 claims description 14
- 230000003111 delayed effect Effects 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 7
- 238000007405 data analysis Methods 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 2
- 230000006399 behavior Effects 0.000 description 19
- 238000004364 calculation method Methods 0.000 description 13
- 230000001960 triggered effect Effects 0.000 description 4
- 230000009467 reduction Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000013524 data verification Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000003245 working effect Effects 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
-
- 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Marketing (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Control Of Vending Devices And Auxiliary Devices For Vending Devices (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A kind of fraud measure and system can not establish a dynamic fraud recognition methods and system for solving the problems, such as, including:Data acquisition module, for obtaining Inner cheat metric data and External Funtions metric data；Rule setting module connects data acquisition module, selects and be arranged to cheat measurement rule for carrying out fraud Measure Indexes；Data processing module, concatenate rule setup module are formatted processing for carrying out data pick-up from Inner cheat metric data and External Funtions metric data, and to data；Identification module is cheated, data processing module is connected, for calculating fraud metric；Human-computer interaction module in customer equipment, connection fraud identification module, for transmitting fraud metric and showing on the client device；Wherein, customer equipment is adjusted fraud identification module intrinsic parameter and fraud metric by human-computer interaction module.Fraud measure of the invention and system improve the flexibility of fraud identification.
Description
Technical Field
The invention relates to the technical field of risk identification, in particular to a fraud measurement method and a fraud measurement system.
Background
Nowadays, online transactions become a part of people's daily life, such as mobile phone payment, online loan application and the like, which greatly facilitates people's life and improves efficiency. However, the rapid update of the technology brings many benefits, and also has the risks of increasing fraud and loan, difficulty in loss finding and the like, and brings loss to the internet financial platform.
The financial fraud involves many business links and means are various, also comparatively hidden. Common fraud behaviors of internet financial services mainly comprise counterfeiting identity registration or falsifying identity registration of other people, stealing or falsifying accounts, applying loans exceeding self repayment capacity, malicious delinquent and the like to a plurality of internet financial platforms. The malicious cheater continuously identifies and breaks the vulnerability of the cheating identification platform, the decision of anti-cheating is a dynamic process, and the internet financial institution cannot guarantee to make a set of 'defense' scheme for resisting all risks.
The existing fraud identification methods mostly perform non-same judgment, are difficult to realize better balance between fraud identification rate and false alarm rate, and cannot establish a dynamic fraud identification method and system.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a fraud measurement system, a fraud identification system and a human-computer interaction module in client equipment perform data interaction, and the flexibility of the system is improved.
The invention also provides a fraud measurement method, which is used for calculating fraud measurement values, outputting different fraud identification results through fraud measurement values in different ranges and improving the flexibility of model identification.
The invention also provides a fraud measurement method, which adjusts the training result by introducing a manual auditing and intervening mode for the measurement value at the boundary, combines the machine identification efficiency and the manual auditing precision and improves the flexibility of model identification.
The technical scheme adopted by the invention for solving the technical problem is as follows:
a fraud metric system comprising:
the data acquisition module is used for acquiring internal fraud measurement data and external fraud measurement data;
the rule setting module is connected with the data acquisition module and is used for carrying out fraud measurement index selection and selecting fraud measurement prevention;
the data processing module is connected with the rule setting module, extracts data from the internal fraud measurement data and the external fraud measurement data and formats the data;
a fraud identification module connected to the data processing module, the fraud identification module being configured to calculate a fraud metric value; and
the human-computer interaction module in the client equipment is connected with the fraud identification module and is used for transmitting the fraud metric value and displaying the fraud metric value on the client equipment;
and the client equipment adjusts the parameters and the fraud metric value in the fraud identification module through the man-machine interaction module.
Preferably, the man-machine interaction module and the fraud recognition module perform data interaction through the internet.
The object of the invention is also achieved by a fraud measurement method comprising:
selecting a fraud metric DiA step (2);
a step of obtaining a target event fraud metric value W by a fraud identification method;
when D is presentiWhen the content is more than or equal to 0,
when D is presentiWhen < 0, W is 0;
wherein, wiMeasuring a weight for fraud; i is an integer between 1 and 6;
judging whether the fraud metric value W is within a set threshold range: if the fraud metric value W is within the set threshold value range, the W and the D are comparediOutputting to a man-machine interaction device, and manually setting a fraud metric value W or adjusting a fraud metric index DiFraud measure weight w ofiThen, returning to the previous step;
if the fraud metric value is not in the set threshold range, when the fraud metric value W is smaller than the lower limit value of the threshold range, the identification result is non-fraud: and when the fraud metric value W is larger than the upper limit value of the threshold range, identifying that the fraud result is fraud.
Preferably, the fraud metric weight wiComprises the following steps:
wherein,is a base value.
Preferably, the fraud metric indicator DiRespectively as customer base data D1Client device data D2Customer transaction data D3Customer behavior data D4Blacklist data D5And multi-platform loan data D6：
Wherein,
wherein, a is the number of loan platforms, and n is the coefficient of the loan platforms; b is the number of delayed repayment; m is the delayed repayment month number; c is the number of violations, kkCoefficient of number of violations, ωiand η is a judgment coefficient, wherein η is 0 if the user is in the blacklist, and η is 1 if the user is not in the blacklist.
Preferably, the fraud identification method is: the method comprises a check comparison method, a cross verification method, a strong characteristic screening method, a risk relation verification method and a behavior data analysis method.
Preferably, the customer base data D1Judging whether triggering is carried out or not by a check comparison method and a cross verification method;
client device data D2Judging whether triggering is carried out or not by a check comparison method and a cross verification method;
customer transaction data D3Judging whether triggering is carried out or not by a strong characteristic screening method and a risk relation verification method;
customer behavior data D4Judging whether triggering is carried out or not by a behavior data analysis method;
blacklist data D5Judging whether triggering is carried out or not by a strong characteristic screening method;
multi-platform loan data D6By the windThe risk relationship verification method judges whether triggering is carried out.
Preferably, the method further comprises the following steps: a step of obtaining internal fraud measurement data and external fraud measurement data of the target event, wherein the step selects fraud measurement index DiIs performed before the step (2).
It is preferable that: the internal fraud metric data includes: fraud metric index DiRespectively as customer base data D1Client device data D2Customer transaction data D3Customer behavior data D4(ii) a The external fraud metric data includes: blacklist data D5And multi-platform loan data D6。
It is preferable that: the client basic data at least includes: name, gender, identification card number, address, contact phone number, bank card number, emergency contact name and phone information;
the client device data includes at least: the computer MAC address, the IP address and the position information of the mobile phone access equipment;
the customer transaction data includes at least: the fund transaction history and overdue condition of the client on the platform;
the customer behavior data includes at least: client landing platform conditions, loan conditions, investment preference conditions;
the blacklist data includes at least: blacklist main body certificate number, contact number, overdue pen number, overdue days, and executed condition by court;
the multi-platform loan data includes at least: borrowing and lending the number of strokes and the total amount on different platforms.
The invention has the beneficial effects that: 1. the core of fraud measurement is that a borrower is taken as the core, and fraud measurement, analysis and prediction are carried out on the basis of various information such as basic information, consumption records, behavior records, unhealthy information and the like. 2. By designing a set of human-computer interaction fraud measurement method and system, the flexibility and vitality of fraud identification are improved.
Drawings
FIG. 1 is an architectural diagram of the fraud measurement system of the present invention.
FIG. 2 is a block diagram of a human-computer interaction module in the fraud measurement system of the present invention.
FIG. 3 is a flow chart of a fraud measurement method of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
A fraud measurement system 100 as shown in fig. 1 includes: a fraud identification system and a human-computer interaction module 106 within the client device, wherein the fraud identification system comprises:
the ① data acquisition module 101 acquires internal fraud measurement ① data and external fraud measurement ① data for identifying fraud risks and carrying out fraud measurement, wherein the ① data acquisition mode comprises ① data actively provided by a user, ① data exchanged with other platforms and fraud related ① data acquired through a crawler technology.
The internal fraud metric data includes: the customer basic data comprises information such as name, gender, identification card number, address, contact phone number, bank card number, emergency contact name and phone; the client equipment data refers to fingerprint data of client access equipment, such as data of a computer MAC address, an IP address, position information of mobile phone access equipment and the like; the client transaction data comprises data of fund transaction history, overdue conditions and the like of a client on the platform; and the customer behavior data comprises data of customer landing platform conditions, loan conditions, investment preference conditions and the like.
The external fraud metric data includes: the blacklist data comprises information such as blacklist main body certificate numbers, contact numbers, overdue strokes, overdue days, execution conditions of the court and the like; the multi-platform loan data comprises data such as loan stroke number, total amount and the like on different platforms.
A rule setting module 102, configured to select fraud metric indexes and a fraud identification method; the fraud metric indicators include: customer base data D1Client device data D2Customer transaction data D3Customer behavior data D4Blacklist data D5And multi-platform loan data D6。
Customer base data D1: the method mainly comprises name, gender, identification card number, address, contact phone number, bank card number, emergency contact name and phone data;
client device data D2: mainly fingerprint data of client access equipment, such as a computer MAC address, an IP address and position information of mobile phone access equipment;
customer transaction data D3: mainly provides fund transaction history and overdue condition data of a client on the platform;
customer behavior data D4: mainly comprises the conditions of a client landing platform, the frequency of modifying client data and the data of applying for loan amount;
blacklist data D5: the method mainly comprises the steps of single main body certificate number, contact telephone, overdue strokes, overdue days and law enforcement;
multi-platform loan data D6: mainly borrow and credit the stroke number, total amount data in different platforms.
The data processing module 103: according to the fraud measurement indexes D1-D6, extracting data from the obtained fraud measurement internal and external data, and formatting the data, for example; the male records the field as 0, and the female records the field as 1; the IP addresses are mapped in a format of a few bits.
The fraud identification module 104: calculating a fraud metric value according to a fraud identification algorithm; the specific calculation method of the fraud metric value refers to step S03;
the internet 105 is a medium through which fraud identification systems interact with client devices;
the man-machine interaction module 106 is used for transmitting the fraud measurement result and the fraud data and parameters according to which the fraud measurement result is judged to the client equipment, displaying the information to the client, and enabling the client to adjust the parameters and establish communication with the fraud identification module.
As shown in fig. 2, the human-computer interaction module 106 includes a parameter area, a result area, and a function area:
wherein, the left data of the parameter area is the processed data received by the fraud identification module 104, and the area value is not settable; the initial value displayed by the right weight of the parameter area is the data received by the fraud identification module 104, the area is settable, after readjustment, a 'parameter reset' button is clicked, a new fraud metric value is generated and is transmitted back to the fraud identification module 104; the result area data is the identification data received by the fraud identification module 104, the value can be manually set after manual judgment, after manual setting, a result reset button is clicked, the fraud metric value is transmitted back to the fraud identification module 104, meanwhile, the weight of each data is reversely calculated and displayed in the man-machine interaction module 106.
The fraud measurement system of the invention utilizes the man-machine interaction module to carry out man-machine interaction, thereby improving the flexibility and vitality of fraud identification.
The invention also protects a fraud measurement method, as shown in fig. 3, which at least comprises the following steps:
step S01, obtaining internal and external fraud measurement data of the target event;
the internal fraud metric data includes: the customer basic data comprises information such as name, gender, identification card number, address, contact phone number, bank card number, emergency contact name and phone; the client equipment data refers to fingerprint data of client access equipment, such as data of a computer MAC address, an IP address, position information of mobile phone access equipment and the like; the client transaction data comprises data of fund transaction history, overdue conditions and the like of a client on the platform; and the customer behavior data comprises data of customer landing platform conditions, loan conditions, investment preference conditions and the like.
The external fraud metric data includes: the blacklist data comprises information such as blacklist main body certificate numbers, contact numbers, overdue strokes, overdue days, execution conditions of the court and the like; the multi-platform loan data comprises data such as loan stroke number, total amount and the like on different platforms.
Step S02, fraud metric D is selectediIndexes;
the fraud metric index includes customer base data D1: the method mainly comprises name, gender, identification card number, address, contact telephone, bank card number, emergency contact name, telephone data and the like;
the fraud metric indicator comprises customer device data D2: mainly fingerprint data of client access equipment, such as a computer MAC address, an IP address, position information of mobile phone access equipment and the like;
the fraud metric includes customer transaction data D3: mainly provides data such as fund transaction history, overdue conditions and the like of a client on the platform;
the fraud metric comprises customer behaviour data D4: mainly comprises data such as the condition of a client landing platform, the frequency of modifying client data, the loan application amount and the like;
the fraud metric indicator comprises internal and external blacklist data D5: the method mainly comprises the steps of single main body certificate numbers, contact calls, overdue strokes, overdue days, execution conditions of a court and the like;
the fraud metric includes multi-platform loan data D6: mainly borrow and credit the number of strokes, total amount and current amount on different platformsTotal amount of borrowed money and the like.
Step S03, obtaining a target event fraud metric value by adopting a fraud identification method;
the fraud identification method comprises five methods of M01-M05:
method M01: check comparison method
And carrying out logic corresponding relation check by utilizing the collected multi-dimensional fraud measurement data. For example, assume that the "identity card number" filled by the user is variable a, the "work property" is variable B, the "location area" of the IP address at the time of application is variable C, and the "common call location" of the mobile phone number is variable D. From the dimension of A + B, if the user age is judged to be less than 20 years according to the identity card number filled by the user, the working property of B is a high-level manager, and C displays an address which is not a commercial district. If there is a significant discrepancy between C and D, the fraud metric is considered high in the C + D dimension. The reliability of the data can be verified from more angles in combination with more variables.
Method M02: cross validation method
And carrying out cross verification on the authenticity and the reliability of the user by utilizing the multi-dimensional data. For example, if the contact address of the applicant is a work address, and the work unit address of the applicant is inconsistent with the filled contact address through external data verification, or the communication addresses of the same company filled by different people are inconsistent, the fraud metric value of the user is higher.
Method M03: strong feature screening method
For the variable with higher default risk, the weight of risk factors is increased, such as the number of multi-head loan times, the number of calls in the night on normal working days, the number of users applying for the login address is abnormally high, and the number of institutions demanding in the call records is large.
Method M04: risk relation verification method
The risk of applicant fraud is judged by identifying the applicant's common contacts for fraud. For example, applicant emergency contacts, or secondary risk associated persons are blacklisted, or applicant device data is blacklisted as data having a higher risk of fraud, etc.
Method M05: behavioral data analysis method
The behavior data analysis method analyzes financial behaviors of the user from behavior data such as consumption habits and living habits of the user. For example, the time of log-in, frequent modification of customer details, application of loans beyond their repayment capacity, etc. may be used as variables to screen their fraud risk and make fraud measures.
The fraud metric value W is calculated as follows:
when D is presentiWhen the content is more than or equal to 0,
when D is presentiWhen < 0, W is 0; (2)
wherein, wiMeasures the weight for fraud, dimensionless; i is an integer between 1 and 6, and has no dimension; the fraud metric value W is between 0 and 100, and has no dimension;
the fraud metric weight wiComprises the following steps:
wherein,is a base value, with no dimension. The fraud metric weight wiAnd fraud metric DiIn inverse relation to each other, fraud measure index DiThe greater the difference from the base value, the fraud measure weight wiThe greater the weight occupied, the more significant the impact of this item of data on the degree of fraud.
First, customer basic data D1Is calculated as follows：
Wherein,for customer base data D1The basic value of (1) is dimensionless, and the default value is 10; omega1The number of triggers is dimensionless and is generally a positive integer.
Calculation example:
base value10, the method M01, M02 determines various basic data provided by the user, when the rule is triggered, the value of the item is correspondingly reduced, and the reduction amplitude (i.e. 0.1 in formula 4 can be set) is in direct proportion to the triggering times, for example, one triggering time is carried outTrigger twiceTrigger three timesThe number of triggers is preferably 5-8, e.g. D1<And (3) substituting 0 into the formula (2), outputting a fraud metric value W as 0, and terminating the judgment of the rest trigger rules.
Second, customer's equipment data D2The calculation formula of (a) is as follows:
wherein,for client device data D2The basic value of (1) is dimensionless, and the default value is 10; omega2The number of triggers is dimensionless and is generally a positive integer.
Calculation example:
base value10, the collected client device data is judged through methods M01 and M02, and when the rule is triggered, the value of the item is correspondingly reduced, and the reduction amplitude (namely 0.2 in formula 5 can be set) is in direct proportion to the triggering times, for example, one triggering time is carried outTrigger twiceTrigger three timesThe general trigger rule is defined as 3-5 times, e.g. D2<If 0 is substituted into the formula (2), the fraud metric value is output as 0, and the judgment of the rest trigger rules is terminated.
Third, the customer transaction data D3The calculation formula of (a) is as follows:
wherein,transacting data D for a customer3The basic value of (1) is dimensionless, and the default value is 20; omega3The number of triggers is dimensionless and is generally a positive integer.
Calculation example:
base valueAt 20, the user transaction data is judged by methods M03 and M04, and when the rule is triggered, the value of the item is correspondingly reduced, and the reduction amplitude (i.e. 0.2 in formula 6 can be set) is in direct proportion to the triggering times, for example, one triggering time is carried outTrigger twiceTrigger three timesThe general trigger rule is defined as 3-4 times, e.g. D3<If 0 is substituted into the formula (2), the fraud metric value is output to be 0, and the judgment of the rest trigger rules is terminated.
Fourth, customer behavior data D4The calculation formula of (a) is as follows:
wherein,for customer behavior data D4The basic value of (1) is dimensionless, and the default value is 10; omega4The number of triggers is dimensionless and is generally a positive integer.
Calculation example:
base valueAt 10, the user behavior data is judged by the method M05, and when the rule is triggered, the value of the item is correspondingly decreased, and the decrease amplitude (i.e. 0.1 in formula 7 can be set) is proportional to the number of triggering, for example, one triggeringTrigger twiceTrigger three timesThe trigger rule is generally defined as 5-8 times, e.g. D4<If 0 is substituted into the formula (2), the fraud metric value is output to be 0, and the judgment of the rest trigger rules is terminated.
Fifth, blacklist data D5The calculation formula of (a) is as follows:
wherein,as blacklist data D5the default value of the basic value is 30, η is a judgment coefficient, if the user is in a blacklist, η is 0, and if the user is not in the blacklist, η is 1.
Calculation example:
base valueAnd if the number is 30, judging the internal and external blacklist data through a method M03, and if the user is in the blacklist, directly setting the value of the item to be 0.
Six, multi-platform loan data D6The calculation formula of (a) is as follows:
wherein,lending data D for multiple platforms6The basic value of (1) is dimensionless, and the default value is 10; a is the number of loan platforms without dimension; n is a loan platform coefficient, has no dimension and generally takes the value of 0.2; b is the number of delayed repayment, has no dimension and is a positive integer; m is the delayed repayment month number, has no dimension and is a positive integer; c is the number of default times without dimension; k is the default number coefficient, has no dimension, and takes the value of 5.
Calculation example:
a base value ofAt 10, the collected multi-platform hospitality data is judged by method M04, and when there is a multi-platform loan without default,when there are multiple platform loans, but there are delayed repayment and the loan is cleared,(m is rounded up, for example, the delayed repayment month number is 3.01 in terms of delayed days/30, and the delayed repayment month number is 4); when there is a multi-platform loan, and there is a default,such as D6<And if 0 is substituted into the formula (2), outputting a fraud metric value of 0, and terminating the judgment of the rest trigger rules.
Step S04, judging whether the fraud metric value W is in the set threshold range;
in this step, the fraud metric value threshold is set to (v)1，v2) Interval (e.g. v)1＝40，v265) less than v1Is determined to be non-fraudulent, is greater than v2Is determined to be fraudulent, the re-calculation or setting of the metric value is performed as a manual intervention during the threshold period.
Step S05, if the fraud metric value W is in the set threshold value range, the measurement result and the index are output to the man-machine interaction equipment;
in this step, the following two processing methods can be selected:
the first method is as follows: adjusting the weights of different measurement indexes, repeating the step S04, and calculating and comparing the fraud measurement value again;
the second method comprises the following steps: and manually setting a fraud metric value, returning to the step S04, and outputting the result to the step S06 after judgment.
Step S06, if the fraud metric value is not within the set threshold range, outputting a fraud identification result: fraud, non-fraud.
While embodiments of the invention have been disclosed above, it is not intended to be limited to the uses set forth in the specification and examples. It can be applied to all kinds of fields suitable for the present invention. Additional modifications will readily occur to those skilled in the art. It is therefore intended that the invention not be limited to the exact details and illustrations described and illustrated herein, but fall within the scope of the appended claims and equivalents thereof.
Claims (10)
1. A fraud metric system, comprising:
the data acquisition module is used for acquiring internal fraud measurement data and external fraud measurement data;
the rule setting module is connected with the data acquisition module and is used for selecting fraud measurement indexes and setting fraud measurement rules;
the data processing module is connected with the rule setting module, extracts data from the internal fraud measurement data and the external fraud measurement data and formats the data;
a fraud identification module connected to the data processing module, the fraud identification module being configured to calculate a fraud metric value; and
the human-computer interaction module in the client equipment is connected with the fraud identification module and is used for transmitting the fraud metric value and displaying the fraud metric value on the client equipment;
and the client equipment adjusts the parameters and the fraud metric value in the fraud identification module through the man-machine interaction module.
2. The fraud metric system of claim 1, wherein the human interaction module and fraud identification module interact data over the internet.
3. A fraud measurement method, comprising:
selecting a fraud metric DiA step (2);
a step of obtaining a target event fraud metric value W by a fraud identification method;
when D is presentiWhen the content is more than or equal to 0,
when D is presentiWhen < 0, W is 0;
wherein, wiMeasuring a weight for fraud; i is an integer between 1 and 6;
judging whether the fraud metric value W is within a set threshold range: if the fraud metric value W is within the set threshold value range, the W and the D are comparediOutputting to a man-machine interaction device, and manually setting a fraud metric value W or adjusting a fraud metric index DiFraud measure weight w ofiThen, returning to the previous step;
if the fraud metric value is not in the set threshold range, when the fraud metric value W is smaller than the lower limit value of the threshold range, the identification result is non-fraud: and when the fraud metric value W is larger than the upper limit value of the threshold range, identifying that the fraud result is fraud.
4. The fraud metric method of claim 3, wherein the fraud metric weight wiComprises the following steps:
wherein,is a base value.
5. The fraud measurement method of claim 3, wherein the fraud measurement indicator DiRespectively as customer base data D1Client device data D2Customer transaction data D3Customer behavior data D4Blacklist data D5And multi-platform loan data D6：
Wherein,
wherein, a is the number of loan platforms, and n is the coefficient of the loan platforms; b is the number of delayed repayment; m is the delayed repayment month number; c is the number of violations, kkCoefficient of number of violations, ωiand η is a judgment coefficient, wherein η is 0 if the user is in the blacklist, and η is 1 if the user is not in the blacklist.
6. The fraud measurement method of claim 5, wherein said fraud identification method is: the method comprises a check comparison method, a cross verification method, a strong characteristic screening method, a risk relation verification method and a behavior data analysis method.
7. The fraud measurement method of claim 6, wherein:
customer base data D1Judging whether triggering is carried out or not by a check comparison method and a cross verification method;
client device data D2Judging whether triggering is carried out or not by a check comparison method and a cross verification method;
customer transaction data D3Judging whether triggering is carried out or not by a strong characteristic screening method and a risk relation verification method;
customer behavior data D4Judging whether triggering is carried out or not by a behavior data analysis method;
blacklist data D5Judging whether triggering is carried out or not by a strong characteristic screening method;
multi-platform loan data D6And judging whether to trigger or not by a risk relation verification method.
8. The fraud measurement method of claim 5 or 7, further comprising:
a step of obtaining internal fraud measurement data and external fraud measurement data of the target event, wherein the step selects fraud measurement index DiIs performed before the step (2).
9. The fraud metric method of claim 8, wherein:
the internal fraud metric data includes: customer base data D1Client device data D2Customer transaction data D3Customer behavior data D4；
The external fraud metric data includes: blacklist data D5And multi-platform loan data D6。
10. The fraud measurement method of claim 9, wherein:
the client basic data at least includes: name, gender, identification card number, address, contact phone number, bank card number, emergency contact name and phone information;
the client device data includes at least: the computer MAC address, the IP address and the position information of the mobile phone access equipment;
the customer transaction data includes at least: the fund transaction history and overdue condition of the client on the platform;
the customer behavior data includes at least: client landing platform conditions, loan conditions, investment preference conditions;
the blacklist data includes at least: blacklist main body certificate number, contact number, overdue pen number, overdue days, and executed condition by court;
the multi-platform loan data includes at least: borrowing and lending the number of strokes and the total amount on different platforms.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810353021.XA CN108846737B (en) | 2018-04-19 | 2018-04-19 | Fraud measurement method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810353021.XA CN108846737B (en) | 2018-04-19 | 2018-04-19 | Fraud measurement method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108846737A true CN108846737A (en) | 2018-11-20 |
| CN108846737B CN108846737B (en) | 2021-03-23 |
Family
ID=64212200
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810353021.XA Active CN108846737B (en) | 2018-04-19 | 2018-04-19 | Fraud measurement method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108846737B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109816519A (en) * | 2019-01-25 | 2019-05-28 | 宜人恒业科技发展（北京）有限公司 | A kind of recognition methods of fraud clique, device and equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102034209A (en) * | 2010-12-15 | 2011-04-27 | 中国民生银行股份有限公司 | Bank credit management method and system |
| CN103714479A (en) * | 2012-10-09 | 2014-04-09 | 四川欧润特软件科技有限公司 | Intelligent centralized monitor method and system for bank personal business fraudulent conducts |
| CN106296343A (en) * | 2016-08-01 | 2017-01-04 | 王四春 | A kind of e-commerce transaction monitoring method based on the Internet and big data |
| CN106506454A (en) * | 2016-10-10 | 2017-03-15 | 江苏通付盾科技有限公司 | Fraud business recognition method and device |
| CN107392451A (en) * | 2017-07-11 | 2017-11-24 | 重庆卡西匚匚科技有限公司 | A kind of risk control system |
| CN107392755A (en) * | 2017-07-07 | 2017-11-24 | 南京甄视智能科技有限公司 | Credit risk merges appraisal procedure and system |
-
2018
- 2018-04-19 CN CN201810353021.XA patent/CN108846737B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102034209A (en) * | 2010-12-15 | 2011-04-27 | 中国民生银行股份有限公司 | Bank credit management method and system |
| CN103714479A (en) * | 2012-10-09 | 2014-04-09 | 四川欧润特软件科技有限公司 | Intelligent centralized monitor method and system for bank personal business fraudulent conducts |
| CN106296343A (en) * | 2016-08-01 | 2017-01-04 | 王四春 | A kind of e-commerce transaction monitoring method based on the Internet and big data |
| CN106506454A (en) * | 2016-10-10 | 2017-03-15 | 江苏通付盾科技有限公司 | Fraud business recognition method and device |
| CN107392755A (en) * | 2017-07-07 | 2017-11-24 | 南京甄视智能科技有限公司 | Credit risk merges appraisal procedure and system |
| CN107392451A (en) * | 2017-07-11 | 2017-11-24 | 重庆卡西匚匚科技有限公司 | A kind of risk control system |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109816519A (en) * | 2019-01-25 | 2019-05-28 | 宜人恒业科技发展（北京）有限公司 | A kind of recognition methods of fraud clique, device and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108846737B (en) | 2021-03-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109919624B (en) | Network loan fraud group recognition and early warning method based on space-time aggregation | |
| US10565592B2 (en) | Risk analysis of money transfer transactions | |
| US8458069B2 (en) | Systems and methods for adaptive identification of sources of fraud | |
| TWI584215B (en) | Method of monitoring suspicious transactions | |
| CN108629593A (en) | Fraudulent trading recognition methods, system and storage medium based on deep learning | |
| US20020194122A1 (en) | Credit extension process using a prepaid card | |
| CN109063977B (en) | Non-inductive transaction risk monitoring method and device | |
| US8117097B2 (en) | Method and system for identifying fraudulent account activity | |
| CN105894380A (en) | Digital currency credit investigation system and method | |
| CN108230151A (en) | A kind of suspicious transaction detection method, apparatus, equipment and storage medium | |
| CN111461704B (en) | Method and system for magnetic stripe card transaction security lock of rural financial institution | |
| CN107644098A (en) | Method, device, equipment and storage medium for identifying fraudulent behavior | |
| CN112365341A (en) | Credit agency anti-fraud method, apparatus, device and storage medium | |
| CN116823274B (en) | Fraud risk assessment method and device applied to financial terminal | |
| KR102005733B1 (en) | Block chain-based person-to-person financial service offering system using credit rating assessment result drawn on online big data analysis | |
| CN112232950A (en) | Loan risk assessment method and device, equipment and computer-readable storage medium | |
| CN108846737B (en) | Fraud measurement method | |
| CN109496317A (en) | The counting implementation method of paper money counter, paper money counter | |
| CN109727116A (en) | Credit analysis method, device, equipment and computer readable storage medium | |
| US20200349600A1 (en) | Customized graphical user interface leveraging dynamic customer relationship chains | |
| CN113362156A (en) | Financial fraud detection and identification system based on Internet of things | |
| CN106570531A (en) | Payment device similarity calculation method and payment device similarity calculation system | |
| CN112967062A (en) | User identity recognition method based on cautious degree | |
| Lonkar et al. | Tackling digital payment frauds: a study of consumer preparedness in India | |
| TW202133086A (en) | Funding demand forecasting method and system |
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 | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A fraud measurement method Effective date of registration: 20230302 Granted publication date: 20210323 Pledgee: Agricultural Bank of China Limited Changchun Chaoyang sub branch Pledgor: CHANGCHUN WHY-E SCIENCE AND TECHNOLOGY Co.,Ltd. Registration number: Y2023220000012 |