CN107305662A - Recognize the method and device of violation account - Google Patents
Recognize the method and device of violation account Download PDFInfo
- Publication number
- CN107305662A CN107305662A CN201610262779.3A CN201610262779A CN107305662A CN 107305662 A CN107305662 A CN 107305662A CN 201610262779 A CN201610262779 A CN 201610262779A CN 107305662 A CN107305662 A CN 107305662A
- Authority
- CN
- China
- Prior art keywords
- value
- affairs
- radix
- broad sense
- numerical value
- 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.)
- Pending
Links
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
- G06Q20/4014—Identity check for transactions
Abstract
This application discloses a kind of method and device for recognizing violation account, methods described includes：Obtain multiple affairs numerical value corresponding with default account；The affairs numerical value is that the default account is produced when carrying out fund transfer affairs with other accounts；Determine the broad sense greatest common divisor of the multiple affairs numerical value；According to the broad sense greatest common divisor and the affairs numerical value, it is determined that confidence value corresponding with the broad sense greatest common divisor；The confidence value characterizes the credibility of the broad sense greatest common divisor；Whether according to the comparative result of the confidence value and predetermined threshold value, it is violation account to judge the default account.Pass through said process so that the efficiency of the violation account in identification network is lifted.
Description
Technical field
The application is related to big data field, more particularly to a kind of method and device for recognizing violation account.
Background technology
At present, the account in network can participate in various forms of fund transfer affairs.Due to shifting thing in fund of realizing
During business, some accounts there may be unlawful practice (such as：Participate in gambling, arbitrage), the account for participating in unlawful practice claims
For violation account.The unlawful practice that general violation account is participated in can produce deleterious effect to platform, therefore need to violation account
It is identified.
In the prior art, the mode of identification violation account typically can be：By the report of related personnel, and to being reported
Account verified (such as：Verified using the historical behavior data of the account reported), it is final to determine to be reported
Account whether be violation account.As can be seen that the mode of prior art identification violation account relatively relies on artificial report, net is recognized
Violation account in network it is less efficient.
The content of the invention
The purpose of the embodiment of the present application is to provide a kind of method and device for recognizing violation account, is deposited with solving prior art
Identification network in violation account it is less efficient the problem of.
The embodiment of the present application uses following technical proposals：
A kind of method for recognizing violation account, it is characterised in that including：
Obtain multiple affairs numerical value corresponding with default account；The affairs numerical value is the default account and other accounts
Carry out fund shifts what is produced during affairs；
Determine the broad sense greatest common divisor of the multiple affairs numerical value；
According to the broad sense greatest common divisor and the affairs numerical value, it is determined that corresponding with the broad sense greatest common divisor can
Certainty value；The confidence value characterizes the credibility of the broad sense greatest common divisor；
Whether according to the comparative result of the confidence value and predetermined threshold value, it is violation account to judge the default account.
A kind of device for recognizing violation account, including：
Numerical value acquiring unit, for obtaining multiple affairs numerical value corresponding with default account；The affairs numerical value is described
What default account was produced when carrying out fund transfer affairs with other accounts；
Broad sense greatest common divisor determining unit, the broad sense greatest common divisor for determining the multiple affairs numerical value；
Confidence value determining unit, for according to the broad sense greatest common divisor and the affairs numerical value, it is determined that with it is described
The corresponding confidence value of broad sense greatest common divisor；The confidence value characterizes the credibility of the broad sense greatest common divisor；
Identifying unit, for the comparative result according to the confidence value and predetermined threshold value, judges that the default account is
No is violation account.
At least one above-mentioned technical scheme that the embodiment of the present application is used can reach following beneficial effect：
The technical scheme provided from each embodiment of above the application, the application utilizes violation account (such as：Participate in gambling
Account) carry out fund shift affairs be produce affairs numerical value have stronger multiple proportion rule, to violation account
Family is set.Specifically, by obtaining multiple affairs numerical value corresponding with default account, and the multiple affairs numerical value is determined
Broad sense greatest common divisor, and determine to the confidence value for the degree of reliability for characterizing the broad sense greatest common divisor, finally, according to
The comparative result of the confidence value and predetermined threshold value, judges whether the default account is violation account.By said process, with
Multiple affairs numerical value that default account is produced when carry out fund shifts affairs are foundation, are realized to the violation account in network
Accurately identify so that the recognition efficiency of violation account is lifted.
Brief description of the drawings
Accompanying drawing described herein is used for providing further understanding of the present application, constitutes the part of the application, this Shen
Schematic description and description please is used to explain the application, does not constitute the improper restriction to the application.In the accompanying drawings：
The flow chart of the method for the identification violation account that Fig. 1 provides for the embodiment of the application one；
Fig. 2 is the result schematic diagram that the application is fitted to one group of affairs numerical value of example；
The structural representation of the device for the identification violation account that Fig. 3 provides for the embodiment of the application one.
Embodiment
To make the purpose, technical scheme and advantage of the application clearer, below in conjunction with the application specific embodiment and
Technical scheme is clearly and completely described corresponding accompanying drawing.Obviously, described embodiment is only the application one
Section Example, rather than whole embodiments.Based on the embodiment in the application, those of ordinary skill in the art are not doing
Go out the every other embodiment obtained under the premise of creative work, belong to the scope of the application protection.
Below in conjunction with accompanying drawing, the technical scheme that each embodiment of the application is provided is described in detail.
Embodiment 1
The embodiment of the present application provides a kind of method for recognizing violation account, in the identification network to solve prior art presence
Violation account it is less efficient the problem of.
The embodiment of the present application provide identification violation account method executive agent can be mobile phone, tablet personal computer and
At least one of terminal devices such as PC (Personal Computer, PC).In addition, the executive agent of this method,
The application program (Application, APP) on terminal device can be mounted in, or can also be server.Described holds
Row main body does not constitute the restriction to the application.
The flow that implements of this method is referred to shown in Fig. 1, mainly comprise the following steps：
Step S101：Obtain multiple affairs numerical value corresponding with default account；The affairs numerical value is the default account
Produced when carrying out fund transfer affairs with other accounts.
Above-mentioned fund transfer affairs can include trade transactions, affairs of transferring accounts carried out based on internet business platform etc.,
The trade transactions typically can jointly be completed by buyer's account and seller's account.Above-mentioned default account can be seller's account or
Buyer's account.If above-mentioned fund transfer affairs are trade transactions, above-mentioned affairs numerical value can be default account (seller) and its
The turnover of the every transaction event carried out between his account (buyer).Above-mentioned affairs numerical value can include：Per transaction event
The amount of money, or bargain quantity per transaction event, etc..
For example, wherein, for participate in gambling activities businessman for, the commodity that the businessman sells can be game money
Deng gambling chips, and gambling chips typically have a unit price, and it is pre- that businessman, which can set the quantity purchases of these gambling chips,
Fixed is several, so, and above-mentioned affairs numerical value can also be the quantity of the gambling chips of merchant transaction, for example, certain participates in gambling
The quantity that businessman can set user's purchase chip must be multiple (100 gambling chips are started at) with 100, then, user a can be with
Purchase 1000 gambling chips, user b can buy 5000 gambling chips etc..
Step S102：Determine the broad sense greatest common divisor of the multiple affairs numerical value.
Herein the technical program is introduced so that affairs numerical value is the amount of money per transaction affairs as an example.Treated for any one
For the default account of identification, during the default account carries out default affairs, can constantly it produce by multiple number of transactions
It is worth the numerical value set constituted.Usually, numerical value set corresponding with each default account can be stored in by number by server
, so, just can be corresponding with the default account of certain to be identified by periodically or non-periodically being obtained from database according in storehouse
Numerical value set, and be identified according to each affairs numerical value (amount of money) in the numerical value set to preset account to this.It is worth mentioning
, to reduce the consumption to computer hardware resource, can intercept above-mentioned default account some period (such as：1 day) in
The affairs numerical value of generation, the foundation recognized as violation account.
In addition, the accuracy to improve the identification of violation account, generally requires and is wrapped to presetting in the numerical value set that account is produced
Each affairs numerical value contained is pre-processed.The pretreatment includes but is not limited to：By affairs numerical value according to monetary exchange rate, it is converted into
Unified monetary unit.Such as, all number of transactions primary systems one in numerical value set are converted into RMB (RMB) unit.The following is
Exemplary numerical value set after being pre-processed：
X={ 4.9,4.8,9.85,9.8,10,20,19.9,19.8,49.5,99,99,99.5 }.
As previously mentioned generally, if multiple affairs numerical value corresponding with some default account have very strong multiple proportion,
In other words, there is the stronger trend using some numerical value as broad sense greatest common divisor (will be described below), then can be assumed that
The default account once or participates in unlawful practice (such as：Gambling, arbitrage behavior etc.).The method of the embodiment of the present application is just
It is according to above-mentioned rule, to recognize the violation account for participating in unlawful practice.It is noted that rule (the broad sense highest common divisor
Number) can be a link for recognizing violation account, during actually identification violation account, other can also be included
Multiple links are (such as：It is identified according to the creditworthiness of seller, social networks network etc.) investigated layer by layer, with final accurate
Violation account is determined, and carries out corresponding punitive measures (such as title).
Greatest common factor (also referred to as greatest common divisor, greatest common factor (G.C.F.)), refers to that two or more integers are had in approximate number most
Big one.For example：50th, 100,150, the greatest common divisor of three is 50.However, the number of transactions produced in network trading scene
Value is not often integer, but real number is (such as：99.8,199.5,299), thus can not according to greatest common divisor method for solving
Calculate the greatest common divisor for obtaining multiple real numbers.In this regard, the embodiment of the present application proposes broad sense greatest common divisor, it is for multiple
Real number, calculates according to certain method and obtains a matching degree highest broad sense greatest common divisor (can be real number).For example：
99.8,199.5,299, the broad sense greatest common divisor of three can be about 100.If being described with mathematical linguistics, for a reality
Manifold S={ x1, x2..., xn, its ε-maximum broad sense common divisor under p norms：
In the embodiment of the present application, to realize the unlawful practice identification to presetting account, it is thus necessary to determine that the thing in numerical value set
The broad sense greatest common divisor of business numerical value, and calculate confidence value corresponding with the broad sense greatest common divisor.Assuming that certain set of values
Close X={ x1, x2..., xnClosed for the amount of money manifold of n transaction of seller's account, then before above-mentioned steps 101, it can determine in advance
An adopted object function：
Wherein,To utilize window function Φ (j μ-xi) build the object function
Obj's is several again；Described times several：For reflecting the affairs numerical value：{x1, x2...,
xnUsing radix μ as the matching degree of broad sense greatest common divisor；Described times of several independent variable includes radix μ, j-th and window letter
The weights omega of the corresponding window of numberj, 1≤j≤k, k is the quantity of window；ω=(ω1, ω2..., ωk) be window function Φ (j μ-
xi) weight vectors.
β||ω||2For the weights omega for window corresponding with window function of the object function ObjjPenalty term.Its
In, β is control weight ωjPenalty coefficient (be generally not desirable to weights omegajValue it is excessive).
The reward for radix μ that-λ μ are the object function Obj.Wherein, λ is the award coefficient of radix μ values
(general radix μ value is bigger, more meets the requirement of broad sense greatest common divisor).
In above-mentioned formula (1), { μ, ω1, ω2..., ωkCan need by treating that optimal method is trained
Parameter, and { k, β, λ ..., Φ } can be the parameter by artificial default settings.Typically can be according to specific business demand or warp
Test value and set above-mentioned { k, β, λ ..., Φ }.
What deserves to be explained is, in the application other embodiment, above-mentioned object function Obj can be only comprising several again：And not comprising above-mentioned penalty term and reward item.Or, above-mentioned object function Obj can be wrapped
Containing several again：And one in above-mentioned penalty term and reward item.Certainly, according to reality
Business demand, can also build other items in above-mentioned object function Obj, and the application is not restricted to this.In addition, on above-mentioned
It is several again, in addition to being built by window function, it can also be built by the function of other forms, to this herein no longer
Enumerated one by one.
In addition what deserves to be explained is, in the embodiment of the application one, the independent variable of the object function Obj can include base
Number μ, the weights omega of j-th of window corresponding with window functionj, etc..However, in other embodiments, above-mentioned object function Obj
Independent variable can be not limited to above parameter.Similarly, what is included in the object function Obj of above-mentioned structure treats that training is obtained
Parameter be also not limited to { μ, ω1, ω2..., ωk, by manually setting the need for including in the object function Obj of above-mentioned structure
Fixed parameter is also not limited to { k, β, λ ..., Φ }.
It should be noted that above-mentioned window function can be the combination of any one or more in following window functions：
Rectangular window：
Quarter window：
Hanning window：
Gaussian window：
In addition, above-mentioned window function includes but is not limited to：Chebyshev window, Kaiser windows, Papoulis windows, Parzen windows,
Poisson windows, Cauchy windows, Gaussian windows, Bartlett-Hann, Blackman-Harris, Nuttall's
Blackman-Harris、Bohman、Flat Top window、Hann、Parzen(de la Valle-Poussin)、
Tapered cosine etc..
In the embodiment of the present application, by taking Gaussian window as an example, the object function Obj in above-mentioned formula (1) is：
Then, based on above-mentioned formula (2), gradient function corresponding with the radix μ can be obtained with derivation：
With the weights omegajCorresponding gradient function：
By above-mentioned formula (3) and formula (4), can use optimization algorithm solve obtain object function Obj because becoming
The value of μ when measuring minimum (μ value is broad sense greatest common divisor).Above-mentioned optimization algorithm can be that gradient declines
Method or L-BFGS (Limited-memory BFGS) method etc..Certainly, in other embodiments, if object function Obj is：
Optimization algorithm can be then used to solve the value for obtaining μ when object function Obj dependent variable takes maximum, and
The value of the μ is defined as broad sense greatest common divisor.
In the embodiment of the present application, above-mentioned formula (3) and formula (4) are obtained in derivation, can be calculated as follows
To broad sense greatest common divisor：
A) determine multiple with 1~μ of the corresponding initial value μ of radix μm, and determine and the weights omegajIt is corresponding initial
Value.
Because above-mentioned object function Obj is not usually convex function, there is local minimum, therefore can be multiple using disposing
Initial value μ corresponding with radix1~μmMethod be trained.On the other hand, each weight corresponding with window function can also be set
Fixed corresponding initial value.For example：
……
Therein, it can be seen that μ1~μmFor 1~m.
B) based on gradient function corresponding with the radix μ and with the weights omegajCorresponding gradient function, using optimal
Change algorithm, calculate the initial value μ obtained with each radix μ1~μmCorresponding radix optimal value μs1~μsm。
As described above, based on formula (3) and formula (4), it is possible to use optimization algorithm (such as gradient descent method or L-
BFGS (Limited-memory BFGS) method), the initial value μ obtained with each radix μ can be calculated1~μmCorresponding radix is excellent
Change value μs1~μsm, for example：
ForIt is 1.05 finally to train obtained μ s1；
ForIt is 2.1 finally to train obtained μ s2；
……
ForFinally train obtained μsmFor (m-0.04)；
Wherein, obtained radix optimal value μ is trained for eachs1~μsm, can correspondingly calculate and obtain weights omega1~
ωkValue.
C) from above-mentioned radix optimal value μs1~μsmIt is middle to determine a value, it is used as the value of the dependent variable in the object function
The value of radix μ during for extreme value.Herein, the extreme value can refer to minimum or maximum.
In the embodiment of the present application, from m radix optimal value μ aboves1~μsmIn, it can select to make object function Obj's
The value of dependent variable is minimum one, is used as global most radix (i.e. broad sense greatest common divisor).As described above, in certain feelings
Condition, can also select to make the value of object function Obj dependent variable be maximum one, be used as global most radix.
It is noted that in the application other embodiment, the process for solving above-mentioned broad sense greatest common divisor can also
It is not limited to above-mentioned setting initial value μ1~μmAnd weights omegajInitial value mode.In addition, in the method for setting initial value,
Set initial value is also not limited to above-mentioned example, for example, can be by weights omegajIt is set as 1/2k.
Fig. 2 is the result schematic diagram that the application is fitted to one group of affairs numerical value of example.It is to be identified in the Fig. 2
Numerical value set in the affairs numerical value that includes be：
X={ 4.9,4.8,9.85,9.8,10,20,19.9,19.8,49.5,99,99,99.5 }.
Wherein, in fig. 2, abscissa represents the value of affairs numerical value, and ordinate represents times in above-mentioned object function Obj
It is severalDependent variable value.It can be seen that：
The position for occurring fluctuating in Fig. 2 is corresponding with the position that affairs numerical value is distributed, for no number of transactions
The position of Distribution value, it is not fluctuated substantially.Wherein, the method introduced according to the above embodiments of the present application, can solve and obtain
The maximum broad sense common divisor of affairs numerical value in above-mentioned numerical value set is：4.9502.
Step S103：According to multiple affairs numerical value in the broad sense greatest common divisor and the numerical value set, it is determined that with
The corresponding confidence value of the broad sense greatest common divisor；The confidence value characterizes the credible journey of the broad sense greatest common divisor
Degree.
Usually, the above method introduced by the embodiment of the present application, can solve and obtain any one numerical value set
Broad sense greatest common divisor, still, different numerical value set, wherein each affairs numerical value, the broad sense obtained with required solution most grand duke
The degree of approximate number into multiple proportion is also different.For example：
Assuming that there is set X={ 4.9,4.8,9.85,9.8,10,20,19.9,19.8,49.5,99,99,99.5 }；Collection
Y={ 3,5.5,9.3,9.8,10.5,22,19,18,48,93,96,102 } is closed,
Assuming that by the above method, the broad sense greatest common divisor that solution obtains above-mentioned set X, Y is 4.9 (about 5), show
So, above-mentioned set X is that to be better than above-mentioned set Y with 4.9 be becoming for broad sense greatest common divisor for the trend of broad sense greatest common divisor with 4.9
Gesture.That is, including the set of multiple affairs numerical value for any one, a broad sense most grand duke may can be solved
Approximate number, but need to characterize the credibility of the broad sense greatest common divisor by corresponding numerical value (above-mentioned confidence value)
(in other words, to characterize the degree of strength into multiple of multiple affairs numerical value and the greatest common divisor in set).
In the embodiment of the application one, determining the process of confidence value can include：
It is determined that the value of the dependent variable of the object function corresponding with each affairs numerical value；
It is determined that the pole of the dependent variable of the object function corresponding with each affairs numerical value, in certain numerical value interval
Value；
It is determined that the dependent variable of the object function corresponding with each affairs numerical value value and the object function because
Ratio between the extreme value of variable；
Determine the average of the ratio；
The average is defined as confidence value corresponding with the broad sense greatest common divisor.
It is if being expressed by formula：
The molecule of above-mentioned formula (5) is meant that：With each affairs numerical value xiIn corresponding, above-mentioned object function Obj
It is several againDependent variable value；The implication of the denominator of above-mentioned formula (5) is in a fixed number
Value is interval (such as：) in the object function dependent variable extreme value.For example, X=4.9,4.8,
9.85,9.8,10,20,19.9,19.8,49.5,99,99,99.5 }, shown in ginseng Fig. 2, for affairs numerical value 9.85, it is assumed that therewith
It is correspondingSuch as：μ=4.9502, then in left and right μ/2 of the affairs numerical value 9.85：2.4751
In the range of, it can findThen：
Then, in this way, above-mentioned ratio can be tried to achieve one by one, and tries to achieve average.In the embodiment of the present application, this is credible
Angle value can be the numerical value between 0~1.
Certainly, in the application other embodiment, determine that the process of above-mentioned confidence value is not limited to the above method.Citing
For：X={ 4.9,4.8,9.85,9.8,10,20,19.9,19.8,49.5,99,99,99.5 }, μ=4.9502；It can pass through
Each affairs numerical value is calculated with solving n times of difference of obtained broad sense greatest common divisor, for 4.9, difference=4.9502-
4.9=0.0502, for 4.8, difference=4.9502-4.8=0.1502, for 9.85, difference=4.9502*2-9.85=
0.0504, by that analogy, can be in the hope of above-mentioned difference corresponding with each affairs numerical value, and solve the difference finally given
Average.
To a certain extent, if above-mentioned difference is smaller (difference is closer to 0), the broad sense greatest common divisor is shown
Credibility it is higher, in other words, if above-mentioned difference is smaller, multiple affairs numerical value in the set and the greatest common divisor
Degree of strength into multiple is higher.
Step S104：According to the comparative result of the confidence value and predetermined threshold value, judge the default account whether as
Violation account.
, can be by the way that above-mentioned confidence value and predetermined threshold value be compared, and according to comparing knot in the embodiment of the present application
Fruit judges whether above-mentioned default account is violation account (as participated in gambling).Specifically, if above-mentioned confidence value is bigger, table
The credibility of the bright broad sense greatest common divisor is higher, then can be by judging whether the confidence value is more than predetermined threshold value
(such as：0.8), come judge the default account whether be violation account (such as：The ＞ 0.8 of confidence value=0.9, judges the default account
Family is violation account, is not otherwise violation account).In another embodiment of the application, if above-mentioned confidence value is smaller, show institute
The credibility for stating broad sense greatest common divisor is higher, then can be by judging whether the confidence value is less than predetermined threshold value (such as：
0.2), come judge the default account whether be violation account (such as：The ＜ 0.2 of confidence value=0.1, judge the default account as
Violation account, is not otherwise violation account).
Based on the above method, the technical scheme provided from each embodiment of above the application, the application utilizes violation account
Family is (such as：Participate in the account of gambling) shift rule of the affairs numerical value with stronger multiple proportion that affairs are generations in carry out fund
Then, violation account set.Specifically, by obtaining multiple affairs numerical value corresponding with default account, and institute is determined
State the broad sense greatest common divisor of multiple affairs numerical value, and determine the credible of the degree of reliability to characterize the broad sense greatest common divisor
Angle value, finally, according to the comparative result of the confidence value and predetermined threshold value, judges whether the default account is violation account.
By said process, to preset multiple affairs numerical value that account produces when carry out fund shifts affairs as foundation, realize to net
Violation account in network is accurately identified so that the recognition efficiency of violation account is lifted.
It is noted that the embodiment of the present application discloses a kind of method that object function is constructed by window function, to ask
The broad sense greatest common divisor of affairs numerical value in skill value set.Certainly, also there are other alternative solutions in the embodiment of the present application.
For example, obfuscation is carried out by the way that the affairs numerical value in logarithm value set is (original), and based on the affairs numerical value after obfuscation,
According to the method for known solution greatest common divisor, and the result that solution is obtained is used as the affairs numerical value in above-mentioned numerical value set
Broad sense greatest common divisor.Wherein, the process of above-mentioned obfuscation may, for example, be rounds up to affairs numerical value.In addition,
The degree of obfuscation can be continuously increased, until finding suitable greatest common divisor.Continuation is illustrated with above-mentioned example：
X={ 4.9,4.8,9.85,9.8,10,20,19.9,19.8,49.5,99,99,99.5 } (original)
Obfuscation is carried out according to rounding-off method, is obtained：
X={ 5,5,10,10,10,20,20,20,50,100,100,100 } (after obfuscation)
So, the affairs numerical value based on above-mentioned obfuscation, it is 5 that can solve and obtain greatest common divisor.
So as to when the affairs numerical value of obfuscation can solve the greatest common divisor for obtaining integer, it is possible to determine that this is preset
Account be violation account (such as：Participate in gambling).It should be noted that herein, above by being solved after obfuscation
To greatest common divisor fall within the category of " broad sense greatest common divisor ".
In another example：
X={ 4.9,4.8,9.85,10.8,10,20,19.9,19.8,49.5,99,99,99.5 } (original)
Obfuscation is carried out according to rounding-off method, is obtained：
X={ 5,5,10,11,10,20,20,20,50,100,100,100 } (after obfuscation)
Then according to the method for known solution greatest common divisor, it is impossible to solve the affairs numerical value after above-mentioned obfuscation most
Big common divisor, can now be given up wherein some affairs numerical value (such as：11), then solved, obtain greatest common divisor
For 5.
In a word, the application solves the mode of the broad sense greatest common divisor of one group of affairs numerical value and does not limit to above-mentioned enumerated
Mode, is no longer introduced one by one herein.
It should be noted that the executive agent that embodiment 1 provides each step of method may each be same equipment, or
Person, this method is also used as executive agent by distinct device.Such as, step S101 and step S102 executive agent can be equipment
1, step S103 executive agent can be equipment 2；Again such as, step S101 executive agent can be equipment 1, step S102
Executive agent with step S103 can be equipment 2；Etc..
Embodiment 2
Whether it is violation account for the default account of identification, the embodiment of the present application provides a kind of device for recognizing violation account,
The structural representation of the device is as shown in figure 3, wherein, the step in function and the above method that the unit in the device is realized
The function of being realized is similar, therefore the device is referred to the content of above method embodiment, is no longer repeated herein.The identification
The device of violation account, can include numerical value acquiring unit 201, broad sense greatest common divisor determining unit 202, confidence value and determine
Unit 203, identifying unit 204.Wherein：
Numerical value acquiring unit 201 is used to obtain multiple affairs numerical value corresponding with default account；The affairs numerical value is institute
State what is produced when default account carries out fund transfer affairs with other accounts.
Broad sense greatest common divisor determining unit 202 is used for the broad sense most grand duke for determining the affairs numerical value in the numerical value set
Approximate number；
Confidence value determining unit 203 is used for the number of transactions in the broad sense greatest common divisor and the numerical value set
Value, it is determined that confidence value corresponding with the broad sense greatest common divisor；The confidence value characterizes the broad sense greatest common divisor
Credibility；
Whether identifying unit 204 is used for according to the confidence value, judge the default account as violation account.
Based on said apparatus, the technical scheme provided from each embodiment of above the application, the application utilizes violation account
Family is (such as：Participate in the account of gambling) shift rule of the affairs numerical value with stronger multiple proportion that affairs are generations in carry out fund
Then, violation account set.Specifically, by obtaining multiple affairs numerical value corresponding with default account, and institute is determined
State the broad sense greatest common divisor of multiple affairs numerical value, and determine the credible of the degree of reliability to characterize the broad sense greatest common divisor
Angle value, finally, according to the comparative result of the confidence value and predetermined threshold value, judges whether the default account is violation account.
By said process, to preset multiple affairs numerical value that account produces when carry out fund shifts affairs as foundation, realize to net
Violation account in network is accurately identified so that the recognition efficiency of violation account is lifted.
In the embodiment of the application one, the broad sense greatest common divisor determining unit 202 can be used for：
According to default object function and the multiple affairs numerical value, it is determined that the value of the dependent variable in the object function
During for extreme value, radix μ value；The independent variable of the object function includes radix μ, and the value of the dependent variable of the object function is anti-
Reflect matching degree of the affairs numerical value using radix μ as broad sense greatest common divisor；
By when the value of the dependent variable of the object function is extreme value, radix μ value is defined as broad sense highest common divisor
Number.
In the embodiment of the present application, described device can also include：
Several construction units, the several again of the object function are built for utilization window function again；Described times of several reflection
Matching degree of the affairs numerical value using radix μ as broad sense greatest common divisor；Described times of several independent variable includes radix μ, jth
The weights omega of individual window corresponding with window functionj, 1≤j≤k, k is the quantity of window.
In the embodiment of the present application, described device can also include one or more of following units：
Penalty term construction unit, the weights omega for window corresponding with window function for building the object functionj
Penalty term；
Item construction unit is rewarded, the reward for radix μ for building the object function.
In the embodiment of the present application, the broad sense greatest common divisor determining unit 202 can include：
Gradient determining unit, for based on the object function, it is determined that gradient function corresponding with the radix μ and with institute
State weights omegajCorresponding gradient function；
Initial value determining unit, for determining multiple initial value μs corresponding with the radix μ1~μm, and determine with it is described
Weights omegajCorresponding initial value；
Highest optimizating unit, for based on gradient function corresponding with the radix μ and with the weights omegajCorresponding gradient
Function, using optimization algorithm, calculates the initial value μ obtained with each radix μ1~μmCorresponding radix optimal value μs1~μsm；
Base value determining unit, for from above-mentioned radix optimal value μs1~μsmIt is middle to determine a value, as in the target
The value of radix μ when the value of the dependent variable of function is extreme value.
In the embodiment of the present application, the optimization algorithm includes gradient descent method or L-BFGS methods.
In the embodiment of the present application, the confidence value determining unit 203 specifically for：
It is determined that the value of the dependent variable of the object function corresponding with each affairs numerical value；
It is determined that the pole of the dependent variable of the object function corresponding with each affairs numerical value, in certain numerical value interval
Value；
It is determined that the dependent variable of the object function corresponding with each affairs numerical value value and the object function because
Ratio between the extreme value of variable；
Determine the average of the ratio；
The average is defined as confidence value corresponding with the broad sense greatest common divisor.
It should be understood by those skilled in the art that, embodiments of the invention can be provided as method, system or computer program
Product.Therefore, the present invention can be using the reality in terms of complete hardware embodiment, complete software embodiment or combination software and hardware
Apply the form of example.Moreover, the present invention can be used in one or more computers for wherein including computer usable program code
The computer program production that usable storage medium is implemented on (including but is not limited to magnetic disk storage, CD-ROM, optical memory etc.)
The form of product.
The present invention is the flow with reference to method according to embodiments of the present invention, equipment (system) and computer program product
Figure and/or block diagram are described.It should be understood that can be by every first-class in computer program instructions implementation process figure and/or block diagram
Journey and/or the flow in square frame and flow chart and/or block diagram and/or the combination of square frame.These computer programs can be provided
The processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that produced by the instruction of computer or the computing device of other programmable data processing devices for real
The device for the function of being specified in present one flow of flow chart or one square frame of multiple flows and/or block diagram or multiple square frames.
These computer program instructions, which may be alternatively stored in, can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that the instruction being stored in the computer-readable memory, which is produced, to be included referring to
Make the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one square frame of block diagram or
The function of being specified in multiple square frames.
These computer program instructions can be also loaded into computer or other programmable data processing devices so that in meter
Series of operation steps is performed on calculation machine or other programmable devices to produce computer implemented processing, thus in computer or
The instruction performed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram one
The step of function of being specified in individual square frame or multiple square frames.
In a typical configuration, computing device includes one or more processors (CPU), input/output interface, net
Network interface and internal memory.
Internal memory potentially includes the volatile memory in computer-readable medium, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only storage (ROM) or flash memory (flash RAM).Internal memory is computer-readable medium
Example.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer-readable instruction, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase transition internal memory (PRAM), static RAM (SRAM), moved
State random access memory (DRAM), other kinds of random access memory (RAM), read-only storage (ROM), electric erasable
Programmable read only memory (EEPROM), fast flash memory bank or other memory techniques, read-only optical disc read-only storage (CD-ROM),
Digital versatile disc (DVD) or other optical storages, magnetic cassette tape, the storage of tape magnetic rigid disk or other magnetic storage apparatus
Or any other non-transmission medium, the information that can be accessed by a computing device available for storage.Define, calculate according to herein
Machine computer-readable recording medium does not include temporary computer readable media (transitory media), such as data-signal and carrier wave of modulation.
It should also be noted that, term " comprising ", "comprising" or its any other variant are intended to nonexcludability
Comprising so that process, method, commodity or equipment including a series of key elements are not only including those key elements, but also wrap
Include other key elements being not expressly set out, or also include for this process, method, commodity or equipment intrinsic want
Element.In the absence of more restrictions, the key element limited by sentence "including a ...", it is not excluded that wanted including described
Also there is other identical element in process, method, commodity or the equipment of element.
It will be understood by those skilled in the art that embodiments herein can be provided as method, system or computer program product.
Therefore, the application can be using the embodiment in terms of complete hardware embodiment, complete software embodiment or combination software and hardware
Form.Deposited moreover, the application can use to can use in one or more computers for wherein including computer usable program code
The shape for the computer program product that storage media is implemented on (including but is not limited to magnetic disk storage, CD-ROM, optical memory etc.)
Formula.
Embodiments herein is the foregoing is only, the application is not limited to.For those skilled in the art
For, the application can have various modifications and variations.It is all any modifications made within spirit herein and principle, equivalent
Replace, improve etc., it should be included within the scope of claims hereof.
Claims (14)
1. a kind of method for recognizing violation account, it is characterised in that including：
Obtain multiple affairs numerical value corresponding with default account；The affairs numerical value is that the default account is carried out with other accounts
Fund shifts what is produced during affairs；
Determine the broad sense greatest common divisor of the multiple affairs numerical value；
According to the broad sense greatest common divisor and the affairs numerical value, it is determined that confidence level corresponding with the broad sense greatest common divisor
Value；The confidence value characterizes the credibility of the broad sense greatest common divisor；
Whether according to the comparative result of the confidence value and predetermined threshold value, it is violation account to judge the default account.
2. according to the method described in claim 1, it is characterised in that determine the broad sense highest common divisor of the multiple affairs numerical value
Number, is specifically included：
According to default object function and the multiple affairs numerical value, it is determined that the value in the dependent variable of the object function is pole
During value, radix μ value；The independent variable of the object function includes radix μ, the value reflection institute of the dependent variable of the object function
State matching degree of the affairs numerical value using radix μ as broad sense greatest common divisor；
By when the value of the dependent variable of the object function is extreme value, radix μ value is defined as broad sense greatest common divisor.
3. method according to claim 2, it is characterised in that it is determined that the broad sense highest common divisor of the multiple affairs numerical value
Before number, methods described also includes：
The several again of the object function are built using window function；The described times of several reflections affairs numerical value using radix μ be wide
The matching degree of adopted greatest common divisor；Described times of several independent variable includes radix μ, j-th window corresponding with window function
Weights omegaj, 1≤j≤k, k is the quantity of window.
4. method according to claim 3, it is characterised in that it is determined that the broad sense highest common divisor of the multiple affairs numerical value
One or more of before number, methods described also comprises the steps：
Build the weights omega for window corresponding with window function of the object functionjPenalty term；
Build the reward for radix μ of the object function.
5. method according to claim 2, it is characterised in that according to default object function and the multiple number of transactions
Value, it is determined that when the value of the dependent variable of the object function is extreme value radix μ value, specifically include：
Based on the object function, it is determined that gradient function corresponding with the radix μ and with the weights omegajCorresponding gradient letter
Number；
Determine multiple initial value μs corresponding with the radix μ1~μm, and determine and the weights omegajCorresponding initial value；
Based on gradient function corresponding with the radix μ and with the weights omegajCorresponding gradient function, using optimization algorithm,
Calculate the initial value μ obtained with each radix μ1~μmCorresponding radix optimal value μs1~μsm；
From above-mentioned radix optimal value μs1~μsmIt is middle to determine a value, during as the value of the dependent variable in the object function for extreme value
Radix μ value.
6. method according to claim 5, it is characterised in that the optimization algorithm includes gradient descent method or L-
BFGS methods.
7. according to the method described in claim 1, it is characterised in that according to the broad sense greatest common divisor and the number of transactions
Value, it is determined that confidence value corresponding with the broad sense greatest common divisor, is specifically included：
It is determined that the value of the dependent variable of the object function corresponding with each affairs numerical value；
It is determined that the extreme value of the dependent variable of the object function corresponding with each affairs numerical value, in certain numerical value interval；
It is determined that the value and the dependent variable of the object function of the dependent variable of the object function corresponding with each affairs numerical value
Extreme value between ratio；
Determine the average of the ratio；
The average is defined as confidence value corresponding with the broad sense greatest common divisor.
8. a kind of device for recognizing violation account, it is characterised in that including：
Numerical value acquiring unit, for obtaining multiple affairs numerical value corresponding with default account；The affairs numerical value is described default
What account was produced when carrying out fund transfer affairs with other accounts；
Broad sense greatest common divisor determining unit, the broad sense greatest common divisor for determining the multiple affairs numerical value；
Confidence value determining unit, for according to the broad sense greatest common divisor and the affairs numerical value, it is determined that with the broad sense
The corresponding confidence value of greatest common divisor；The confidence value characterizes the credibility of the broad sense greatest common divisor；
Identifying unit, for the comparative result according to the confidence value and predetermined threshold value, judge the default account whether as
Violation account.
9. device according to claim 8, it is characterised in that the broad sense greatest common divisor determining unit is used for：
According to default object function and the multiple affairs numerical value, it is determined that the value in the dependent variable of the object function is pole
During value, radix μ value；The independent variable of the object function includes radix μ, the value reflection institute of the dependent variable of the object function
State matching degree of the affairs numerical value using radix μ as broad sense greatest common divisor；
By when the value of the dependent variable of the object function is extreme value, radix μ value is defined as broad sense greatest common divisor.
10. device according to claim 9, it is characterised in that described device also includes：
Several construction units, the several again of the object function are built for utilization window function again；Described in described times of several reflections
Matching degree of the affairs numerical value using radix μ as broad sense greatest common divisor；Described times of several independent variable include radix μ, j-th with
The weights omega of the corresponding window of window functionj, 1≤j≤k, k is the quantity of window.
11. device according to claim 10, it is characterised in that described device also includes one or many in following units
It is individual：
Penalty term construction unit, the weights omega for window corresponding with window function for building the object functionjPunishment
；
Item construction unit is rewarded, the reward for radix μ for building the object function.
12. device according to claim 9, it is characterised in that the broad sense greatest common divisor determining unit includes：
Gradient determining unit, for based on the object function, it is determined that gradient function corresponding with the radix μ and with the power
Weight ωjCorresponding gradient function；
Initial value determining unit, for determining multiple initial value μs corresponding with the radix μ1~μm, and determine and the weight
ωjCorresponding initial value；
Highest optimizating unit, for based on gradient function corresponding with the radix μ and with the weights omegajCorresponding gradient function,
Using optimization algorithm, the initial value μ obtained with each radix μ is calculated1~μmCorresponding radix optimal value μs1~μsm；
Base value determining unit, for from above-mentioned radix optimal value μs1~μsmIt is middle to determine a value, as in the object function
Dependent variable value be extreme value when radix μ value.
13. device according to claim 12, it is characterised in that the optimization algorithm includes gradient descent method or L-
BFGS methods.
14. device according to claim 8, it is characterised in that the confidence value determining unit specifically for：
It is determined that the value of the dependent variable of the object function corresponding with each affairs numerical value；
It is determined that the extreme value of the dependent variable of the object function corresponding with each affairs numerical value, in certain numerical value interval；
It is determined that the value and the dependent variable of the object function of the dependent variable of the object function corresponding with each affairs numerical value
Extreme value between ratio；
Determine the average of the ratio；
The average is defined as confidence value corresponding with the broad sense greatest common divisor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610262779.3A CN107305662A (en) | 2016-04-25 | 2016-04-25 | Recognize the method and device of violation account |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610262779.3A CN107305662A (en) | 2016-04-25 | 2016-04-25 | Recognize the method and device of violation account |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107305662A true CN107305662A (en) | 2017-10-31 |
Family
ID=60150553
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610262779.3A Pending CN107305662A (en) | 2016-04-25 | 2016-04-25 | Recognize the method and device of violation account |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107305662A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112035241A (en) * | 2020-10-12 | 2020-12-04 | 支付宝(杭州)信息技术有限公司 | Task processing method and device |
| CN113050978A (en) * | 2019-12-26 | 2021-06-29 | 中国移动通信集团北京有限公司 | Application gray release control method, device and equipment and computer storage medium |
| CN113436020A (en) * | 2021-08-26 | 2021-09-24 | 深圳华锐金融技术股份有限公司 | Illegal account identification method and device, computer equipment and storage medium |
| CN112035241B (en) * | 2020-10-12 | 2024-05-03 | 支付宝(杭州)信息技术有限公司 | Task processing method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101253507A (en) * | 2005-08-29 | 2008-08-27 | 大金工业株式会社 | Account false use detecting/suppressing device, data collecting device, and account false use detecting/suppressing program |
| CN104199832A (en) * | 2014-08-01 | 2014-12-10 | 西安理工大学 | Financial network unusual transaction community finding method based on information entropy |
| CN104657811A (en) * | 2013-11-18 | 2015-05-27 | 国际商业机器公司 | Method and system for account health assessment, risk identification, and remediation |
| US9064364B2 (en) * | 2003-10-22 | 2015-06-23 | International Business Machines Corporation | Confidential fraud detection system and method |
| CN105427166A (en) * | 2015-11-13 | 2016-03-23 | 中国建设银行股份有限公司 | Bank account type detection method and system |
-
2016
- 2016-04-25 CN CN201610262779.3A patent/CN107305662A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9064364B2 (en) * | 2003-10-22 | 2015-06-23 | International Business Machines Corporation | Confidential fraud detection system and method |
| CN101253507A (en) * | 2005-08-29 | 2008-08-27 | 大金工业株式会社 | Account false use detecting/suppressing device, data collecting device, and account false use detecting/suppressing program |
| CN104657811A (en) * | 2013-11-18 | 2015-05-27 | 国际商业机器公司 | Method and system for account health assessment, risk identification, and remediation |
| CN104199832A (en) * | 2014-08-01 | 2014-12-10 | 西安理工大学 | Financial network unusual transaction community finding method based on information entropy |
| CN105427166A (en) * | 2015-11-13 | 2016-03-23 | 中国建设银行股份有限公司 | Bank account type detection method and system |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113050978A (en) * | 2019-12-26 | 2021-06-29 | 中国移动通信集团北京有限公司 | Application gray release control method, device and equipment and computer storage medium |
| CN113050978B (en) * | 2019-12-26 | 2024-03-19 | 中国移动通信集团北京有限公司 | Gray release control method, device and equipment for application and computer storage medium |
| CN112035241A (en) * | 2020-10-12 | 2020-12-04 | 支付宝(杭州)信息技术有限公司 | Task processing method and device |
| CN112035241B (en) * | 2020-10-12 | 2024-05-03 | 支付宝(杭州)信息技术有限公司 | Task processing method and device |
| CN113436020A (en) * | 2021-08-26 | 2021-09-24 | 深圳华锐金融技术股份有限公司 | Illegal account identification method and device, computer equipment and storage medium |
| CN113436020B (en) * | 2021-08-26 | 2021-12-10 | 深圳华锐金融技术股份有限公司 | Illegal account identification method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI740106B (en) | Product recommendation method and device | |
| TWI712981B (en) | Risk identification model training method, device and server | |
| CN108763277B (en) | Data analysis method, computer readable storage medium and terminal device | |
| US11481687B2 (en) | Machine learning and security classification of user accounts | |
| WO2018149386A1 (en) | Risk management and control method and device | |
| EP3996021A1 (en) | Computer-implemented system and method for generating and extracting user related data stored on a blockchain | |
| CN107437198A (en) | Determine method, information recommendation method and the device of consumer's risk preference | |
| CN107622432A (en) | Trade company's evaluation method and system | |
| CN107679856B (en) | Transaction-based service control method and device | |
| TWI752349B (en) | Risk identification method and device | |
| CN107871276A (en) | Inquiry unit, method and the computer-readable recording medium of loan product | |
| CN108182634A (en) | A kind of training method for borrowing or lending money prediction model, debt-credit Forecasting Methodology and device | |
| CN109766454A (en) | A kind of investor's classification method, device, equipment and medium | |
| CN107203939A (en) | Determine method and device, the computer equipment of consumer's risk grade | |
| CN108961032A (en) | Borrow or lend money processing method, device and server | |
| US20210243215A1 (en) | Fraud detection based on analysis of frequency-domain data | |
| Yuan et al. | The general dynamic risk assessment for the enterprise by the hologram approach in financial technology | |
| CN104951843B (en) | Method for Sales Forecast system and method | |
| CN111708883A (en) | Credit credit limit determination method and device based on machine learning and equipment fingerprint | |
| WO2023064933A1 (en) | A decentralized social news network website application (dapplication) on a blockchain including a newsfeed, nft marketplace, and a content moderation process for vetted content providers | |
| CN108596765A (en) | A kind of Electronic Finance resource recommendation method and device | |
| KR20120032606A (en) | Stock investment system enabling participattion of stock investment clients and method thereof | |
| US8204840B2 (en) | Global conduct score and attribute data utilization pertaining to commercial transactions and page views | |
| CN107305662A (en) | Recognize the method and device of violation account | |
| CN113643060A (en) | Product price prediction method and device |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200921 Address after: Cayman Enterprise Centre, 27 Hospital Road, George Town, Grand Cayman, British Islands Applicant after: Innovative advanced technology Co.,Ltd. Address before: Cayman Enterprise Centre, 27 Hospital Road, George Town, Grand Cayman, British Islands Applicant before: Advanced innovation technology Co.,Ltd. Effective date of registration: 20200921 Address after: Cayman Enterprise Centre, 27 Hospital Road, George Town, Grand Cayman, British Islands Applicant after: Advanced innovation technology Co.,Ltd. Address before: A four-storey 847 mailbox in Grand Cayman Capital Building, British Cayman Islands Applicant before: Alibaba Group Holding Ltd. |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171031 |