CN108898418A - User account detection method, device, computer equipment and storage medium - Google Patents
User account detection method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN108898418A CN108898418A CN201810547162.5A CN201810547162A CN108898418A CN 108898418 A CN108898418 A CN 108898418A CN 201810547162 A CN201810547162 A CN 201810547162A CN 108898418 A CN108898418 A CN 108898418A
- Authority
- CN
- China
- Prior art keywords
- user account
- user
- account
- history
- obtains
- 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
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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0248—Avoiding fraud
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application involves a kind of user account detection method, device, computer equipment and storage mediums.The method includes：User account data are obtained, obtain user characteristics attribute according to user account data；User characteristics attribute is input in pre-set user account number classification device, output feature is obtained；User account testing result is obtained according to output feature.Abnormal user account can effectively be detected using this method.
Description
Technical field
This application involves field of computer technology, set more particularly to a kind of user account detection method, device, computer
Standby and storage medium.
Background technique
In most internet platforms, in order to improve the liveness of platform user, migration efficiency can be set, as registration send it is red
It wraps, discount coupon, and consumption is sent to return and show, return discount coupon, movable discounted price etc..But these movable welfares may be by exception
User account batch operation is got using loophole, is not reached in normal users account hand directly, is brought to platform
Huge economic loss.To prevent from getting welfare by abnormal user account batch or getting welfare using platform loophole, adopt at present
With modes such as various safeguard procedures, such as identifying code, short-message verification code check, but all there is defects for these modes, abnormal
User account user is very easy to obtain welfare around these safeguard procedures, causes largely to lose to internet platform.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide a kind of user that can effectively detect abnormal user account
Account detection method, device, computer equipment and storage medium.
A kind of user account detection method, the method includes：
User account data are obtained, obtain user characteristics attribute according to user account data；
User characteristics attribute is input in pre-set user account number classification device, output feature is obtained；
User account testing result is obtained according to output feature.
The generation step of the pre-set user account number classification device includes in one of the embodiments,：
Historical user's account data and corresponding testing result are obtained, testing result includes history normal users account and goes through
History abnormal user account；
According to historical user's account data and corresponding testing result statistical history user account number, history normal users account
Number and history abnormal user account number, and calculate history normal users account frequency and history abnormal user account frequency；
Corresponding historical user's feature is obtained according to historical user's account data, and to historical user's feature according to default item
Part is divided, and item to be sorted is obtained；
Count the corresponding historical user's account number of item to be sorted, history normal users account number and history abnormal user account
Number, and calculate conditional probability and historical user's account that the corresponding historical user's account of item to be sorted is history normal users account
For the conditional probability of history abnormal user account, pre-set user account number classification device is obtained.
It is described in one of the embodiments, that the user characteristics attribute is input in pre-set user account number classification device,
Output feature is obtained, including：
Obtain the corresponding target item to be sorted of user characteristics attribute；
The corresponding conditional probability of target item to be sorted is obtained, user is calculated separately using Bayes' theorem according to conditional probability
Account is normal users account probability and user account is abnormal user account probability；
Compare normal users account probability and abnormal user account probability, output feature is obtained according to comparison result.
The method in one of the embodiments, further includes：
User account shipping address information is obtained according to user account data；
User account shipping address information is segmented, word segmentation result is obtained, word segmentation result is input to Clustering Model
In obtain classification results, obtain shipping address similarity according to classification results；
Doubtful abnormal user account is obtained according to shipping address similarity；
User account data are then obtained, including：
Obtain the user account data of doubtful abnormal user account.
The described word segmentation result is input in Clustering Model obtains classification results in one of the embodiments, packet
It includes：
User account shipping address information is grouped according to preset condition, total number packets are calculated, according to total number packets
Calculate clusters number；
The target word of clusters number is obtained from word segmentation result as initial cluster center, using initial cluster center as current cluster
Center；
Other words in word segmentation result in addition to target word are obtained, calculate other words in addition to target word into current cluster
The distance of the heart；
Other words in addition to target word are assigned in the corresponding cluster in current cluster center according to distance, obtain clusters number
Target cluster；
Calculate target cluster target cluster center, using target cluster center be used as current cluster center, return calculate remove target word with
Outer other words to current cluster center apart from the step of carry out repeating cluster, when meeting the condition of convergence, obtain classification knot
Fruit.
The method in one of the embodiments, further includes：
Input feature value is obtained according to user account data, input feature value is input to the detection of pre-set user account
In model, output feature vector is obtained；
User account testing result is obtained according to output feature vector.
The generation step of the pre-set user account detection model in one of the embodiments, including：
Historical user's account data and corresponding testing result are obtained, history input is obtained according to historical user's account data
Feature vector obtains history output feature vector according to testing result；
Using history input feature value as the input of Logic Regression Models, returned using history output feature vector as logic
The label of model is returned to be trained；
When the cost function of Logic Regression Models reaches preset threshold, pre-set user account detection model is obtained.
A kind of user account detection device, described device include：
Characteristic attribute obtains module, for obtaining user account data, obtains user characteristics category according to user account data
Property；
Output feature obtains module, for user characteristics attribute to be input in pre-set user account number classification device, obtains defeated
Feature out；
Testing result obtains module, for obtaining user account testing result according to output feature.
A kind of computer equipment can be run on a memory and on a processor including memory, processor and storage
Computer program, the processor realize following steps when executing the computer program：
User account data are obtained, obtain user characteristics attribute according to user account data；
User characteristics attribute is input in pre-set user account number classification device, output feature is obtained；
User account testing result is obtained according to output feature.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
Following steps are realized when row：
User account data are obtained, obtain user characteristics attribute according to user account data；
User characteristics attribute is input in pre-set user account number classification device, output feature is obtained；
User account testing result is obtained according to output feature.
Above-mentioned user account detection method, device, computer equipment and storage medium, by obtaining user account data,
User characteristics attribute is obtained according to user account data；User characteristics attribute is input in pre-set user account number classification device, is obtained
To output feature；User account testing result is obtained according to output feature, the detection of pre-set user account number classification device is able to use and uses
Family account data obtains user account testing result, can effectively detect abnormal user account.
Detailed description of the invention
Fig. 1 is the application scenario diagram of user account detection method in one embodiment；
Fig. 2 is the flow diagram of user account detection method in one embodiment；
Fig. 3 is to obtain the flow diagram of pre-set user account number classification device in one embodiment；
Fig. 4 is the flow diagram that output feature is obtained in one embodiment；
Fig. 5 is to obtain the flow diagram of doubtful abnormal user account in one embodiment；
Fig. 6 is the flow diagram clustered in one embodiment according to user account shipping address；
Fig. 7 is the flow diagram of user account detection method in another embodiment；
Fig. 8 is to obtain the flow diagram of pre-set user account detection model in one embodiment；
Fig. 9 is the structural block diagram of user account detection device in one embodiment；
Figure 10 is the internal structure chart of computer equipment in one embodiment.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
User account detection method provided by the present application, can be applied in application environment as shown in Figure 1.Wherein, eventually
End 102 is communicated with server 104 by network by network.Server 104 gets the user account of the transmission of terminal 102
Data obtain user characteristics attribute according to user account data；User characteristics attribute is input to pre-set user account number classification device
In, obtain output feature；User account testing result is obtained according to output feature.Wherein, terminal 102 can be, but not limited to be each
Kind personal computer, laptop, smart phone, tablet computer and portable wearable device, server 104 can be with solely
The server clusters of the either multiple servers compositions of vertical server is realized.
In one embodiment, as shown in Fig. 2, providing a kind of user account detection method, it is applied to Fig. 1 in this way
In server for be illustrated, include the following steps：
S202 obtains user account data, obtains user characteristics attribute according to user account data.
Wherein, user account data include account primary attribute, facility information, user behavior data and business datum.Its
In, it may include account name, cell-phone number, bank's card number, name, year that account primary attribute, which is used to reflect the personal information of user,
Age, gender, identity card and address etc..Facility information may include mobile phone, flat for describing the device parameter information that user uses
The parameter of the various kinds of equipment such as plate computer, notebook and PC is also possible to the device-fingerprint that equipment is commonly used in user.User behavior
Data refer to the Various types of data that user generates when carrying out various operations on webpage or client, may include that User Page stops
Stay duration, the access order of user, operating frequency and key information etc..The number of generation when business datum refers to progress business activity
According to, for example, there is the second to kill commodity business activity when, then business datum is exactly the dimensional attribute of the commodity.
User characteristics attribute refers to that the corresponding value set of user characteristics, the value set are to user account data by handling
It obtains, user characteristics include primary attribute feature, facility information feature, user behavior characteristics and service feature etc., Yong Hute
Sign is to carry out feature extraction according to historical user's account data to obtain.Wherein, it can be by processing according to user account number
According to carrying out that user characteristics attribute is calculated, it is also possible to pre-set user account data corresponding with user characteristics attribute
Relationship obtains user characteristics attribute, such as：Primary attribute feature a1 isIf good in a user account data
Friendly quantity is 100 good friends, registers number of days as 400 days, then the corresponding user characteristics attribute of user characteristics a1 can calculate
To being 0.25.For another example facility information feature b 1 is device chip model, if device chip model in a user account data
Corresponding relationship for X86 chip, the then basis is pre-set device chip model and user characteristics attribute obtains the user
The corresponding user characteristics attribute of feature b 1 is 1.
Specifically, it in the case where obtaining user's authorization, can go to obtain user's account using different collecting methods
Number can obtain user account data by burying a little in user terminal progress business, can also be by user terminal
Preset data acquires script, starts script in the load of the platform page and obtains user account data, can also by getting
Log information in server obtains user account data from log information.Then data cleansing is carried out to the data got,
Data cleansing refers to discovery and corrects last one of program of identifiable mistake in data file, including checks that data are consistent
Property, handle invalid value and missing values etc..Data characteristics after cleaning is extracted, user characteristics are obtained, according to user account data
The corresponding user characteristics attribute of user characteristics is got, user characteristics attribute is obtained.
User characteristics attribute is input in pre-set user account number classification device by S204, obtains output feature.
Wherein, pre-set user account number classification device makes to train historical user's account data to obtain in advance according to NB Algorithm
It arrives.Output is characterized in the characteristic value for judging testing result, sets not previously according to historical user's account testing result
With testing result, corresponding different output feature, that is, correspond to different characteristic values, for example pre-set testing result and be positive
The normal corresponding output feature, that is, characteristic value of user account is 1, and pre-setting testing result is that abnormal user account is corresponding defeated
Feature, that is, characteristic value is 0 out.The output feature then obtained can be 1 or 0.
Specifically, the user characteristics attribute obtained according to user account data is input to and is calculated previously according to naive Bayesian
In the user account classifier of method training, the corresponding output feature of testing result has been obtained.Wherein, NB Algorithm is base
In the classification method that Bayes' theorem and characteristic condition are independently assumed.
S206 obtains user account testing result according to output feature.
Specifically, it according to the corresponding relationship of preset output feature and historical user's account testing result, obtains defeated at this time
The corresponding user account testing result of feature out.For example, if pre-setting user account testing result is normal users account
When corresponding output feature be 1, when user account testing result is abnormal user account, corresponding output feature is 0.Then when
When output feature is 1, illustrates that user account at this time is normal users account, if output feature is 0, illustrate user at this time
Account is abnormal user account.At this point, can be to the various of the abnormal user account when user account is abnormal user account
Operation requests carry out real-time blocking.
In above-mentioned user account detection method, by obtaining user account data, user is obtained according to user account data
Characteristic attribute；User characteristics attribute is input in pre-set user account number classification device, output feature is obtained；It is obtained according to output feature
To user account testing result.It is able to use pre-set user account number classification device detection user account data and obtains user account detection
As a result, it is possible to effectively detect abnormal user account.It, can be to abnormal user account when detecting as abnormal user account
Various operation requests intercepted, abnormal user account can be effectively reduced to loss of economic benefit caused by platform.
In one embodiment, as shown in figure 3, the generation step of the pre-set user account number classification device includes：
S302, obtains historical user's account data and corresponding testing result, testing result include history normal users account
Number and history abnormal user account.
Wherein, history normal users account refers to the use that not abnormal phenomenon is detected by historical user's account data
Family account is history normal users account.History abnormal user account, which refers to, carries out detection discovery by historical user's account data
The user account of abnormal phenomenon is history abnormal user account.Abnormal phenomenon refers to the operation requests of the user in business activity
The phenomenon that not meeting preset rule of conduct with operation behavior or user account are logged in using warping apparatus, in different-place login etc.
Phenomenon.
Specifically, historical user's account data and corresponding testing result are got, historical user's account data and right
The testing result answered, which can be, carries out the testing result that detection history user account data obtain by artificial or Expert Rules
What then data were saved, using obtained historical user's account data and corresponding testing result as sample data.
S304 is normal according to historical user's account data and corresponding testing result statistical history user account number, history
User account number and history abnormal user account number, and calculate history normal users account frequency and history abnormal user account frequency
Rate.
Specifically, historical user's account number in statistical sample data, history normal users account number and history are used extremely
Family account number.History normal users account frequency is calculated according to historical user's account number and history normal users account number, according to
Historical user's account number and history abnormal user account number calculate history abnormal user account frequency.
S306 obtains corresponding historical user's feature according to historical user's account data, and to historical user's feature according to
Preset condition is divided, and item to be sorted is obtained.
Wherein, item to be sorted, which refers to divide user characteristics according to preset condition, obtains different division results later, often
A user characteristics can correspond to multiple divisions, it can corresponding multiple items to be sorted.Preset condition can be according to artificial warp
Test the division condition of historical user's feature of setting.For example, the historical user's feature extracted can be
Then the item to be sorted of the user characteristics can be a1<=0.05,0.05<a1<0.2 and a1>=0.2.If historical user's account
Daily record data in number and register the ratio between number of days as 0.15, user's head portrait be really be head portrait, then its user characteristics attribute is
0.15, the item to be sorted belonged to is 0.05<a1<0.2.Historical user's feature can be a2=user's head portrait, then the user is special
It is true head portrait or other head portraits that the item to be sorted of sign, which can be user's head portrait,.
Specifically, feature extraction is carried out according to historical user's account data and obtains corresponding historical user's feature, and to mentioning
The each historical user's feature got is divided according to preset condition, obtains the corresponding item to be sorted of each user characteristics.
S308 counts the corresponding historical user's account number of item to be sorted, history normal users account number and history and uses extremely
Family account number, and calculate conditional probability and history use that the corresponding historical user's account of item to be sorted is history normal users account
Family account is the conditional probability of history abnormal user account, obtains pre-set user account number classification device.
Specifically, the historical user's account number, history normal users account number and history for counting each item to be sorted are abnormal
User account number.And calculate the corresponding historical user's account of each item to be sorted be history normal users account conditional probability and
Historical user's account is the conditional probability of history abnormal user account, i.e., calculating historical user's account is history normal users account
Or under conditions of history abnormal user occurs, the probability that corresponding each item to be sorted occurs just has obtained pre-set user account
Classifier.
In one embodiment, sample data can be divided into training sample data and test sample data, use training
Sample data is trained to obtain initial user account number classification device, after obtaining pre-set user account number classification device, uses test
Sample data tests initial user account number classification device, and when test result reaches default accuracy, test is completed, and obtains
Pre-set user account number classification device.When test result does not reach default accuracy, then initial user account number classification is re-started
The training of device, available more training sample data are trained, until test result reaches default accuracy.
In the above-described embodiments, by obtaining historical user's account data and corresponding testing result, the testing result
Including history normal users account and history abnormal user account；It is tied according to historical user's account data and corresponding detection
Fruit statistical history user account number, history normal users account number and history abnormal user account number, and it is just common to calculate history
Family account frequency and history abnormal user account frequency；Corresponding historical user's feature is obtained according to historical user's account data,
And historical user's feature is divided according to preset condition, obtain item to be sorted；It is corresponding to count the item to be sorted
Historical user's account number, history normal users account number and history abnormal user account number, and it is corresponding to calculate the item to be sorted
Historical user's account be the conditional probability of history normal users account and historical user's account is history abnormal user account
Conditional probability obtains pre-set user account number classification device.By pre-setting user account classifier, user account inspection is being carried out
It when survey, can directly use, improve the efficiency of user account detection.
In one embodiment, as shown in figure 4, step S204, i.e., it is described the user characteristics attribute is input to it is default
In user account classifier, output feature, including step are obtained：
S402 obtains the corresponding target item to be sorted of user characteristics attribute.
Specifically, the corresponding target item to be sorted of user characteristics attribute is obtained according to item to be sorted ready-portioned in advance, often
One user characteristics can all have a corresponding target item to be sorted.
S404 is obtained the corresponding conditional probability of target item to be sorted, is counted respectively according to conditional probability using Bayes' theorem
Calculation user account is normal users account probability and user account is abnormal user account probability.
Specifically, the corresponding conditional probability of item to be sorted is got, the conditional probability is according to preparatory trained user's account
What number classifier obtained, user account is calculated according to the conditional probability and history normal users account frequency usage Bayes' theorem
For normal users account probability, user is calculated according to the conditional probability and history abnormal user account frequency usage Bayes' theorem
Account is abnormal user account probability.Wherein, Bayesian calculation formula isWherein, P
Under conditions of (B | A) refers to that item A to be sorted occurs, user account is the probability of normal users account or abnormal user account, P (A
| B) refer to the corresponding conditional probability of item to be sorted, P (B) refers to history normal users account frequency or history abnormal user account
Number frequency.It is then P according to the calculation formula that Bayes' theorem obtains when item to be sorted is multinomial again because denominator is constant
(B | A)=P (A1|B)P(A2|B)...P(Am| B) P (B), wherein AmIndicate the conditional probability of m-th of item to be sorted.
S406 compares normal users account probability and abnormal user account probability, obtains output feature according to comparison result.
Specifically, compare normal users account probability and abnormal user account probability, then export the big conduct of probability special
Sign.For example, the corresponding output feature of abnormal user account is 0, if comparing when the corresponding output feature of normal users account is 1
As a result big for normal users account probability, then the output feature obtained is 1, if comparison result is that abnormal user account probability is big,
The output feature then obtained is 0.
In above-described embodiment, by obtaining the corresponding target item to be sorted of user characteristics attribute；Obtain target item to be sorted
Corresponding conditional probability, according to conditional probability using Bayes' theorem calculate separately user account be normal users account probability and
User account is abnormal user account probability；Compare normal users account probability and abnormal user account probability, is tied according to comparing
Fruit obtains output feature, can be more convenient and be accurately obtained output feature.
In one embodiment, as shown in figure 5, the method, further comprising the steps of, following steps are doubtful for obtaining
Abnormal user account：
S502 obtains user account shipping address information according to user account data.
Wherein, user account shipping address information refers to the shipping address information of user account commodity when being traded.
Specifically, collecting cargo in detail according to what is obtained in the account primary attribute in obtained user account data
Location information, or the better address information of receiving of business activity commodity got from business datum.Different user accounts
There is different shipping addresses.Such as：User account shipping address can be the city XX of the XX province county the XX main road the XX garden XX XX.
S504, segments user account shipping address information, obtains word segmentation result, and word segmentation result is input to cluster
Classification results are obtained in model, obtain shipping address similarity according to classification results.
Wherein, Clustering Model refers to the disaggregated model established using clustering algorithm, and shipping address similarity user describes to receive
The similarity degree of goods address.
Specifically, segmenting to user account shipping address information, word segmentation result is obtained, using word segmentation result as word set
Conjunction, which is input in Clustering Model, obtains classification results, obtains shipping address similarity according to classification results.For example,：Above-mentioned user
Obtained word segmentation result is XX province, the city XX, the county XX, the main road XX, the garden XX and No. XX after account shipping address participle.
S506 obtains doubtful abnormal user account according to shipping address similarity.
Specifically, illustrating the high user's account of these shipping address similarity-rough sets when shipping address similarity-rough set is high
A possibility that number exception, is higher, then using the higher user account of shipping address similarity as doubtful abnormal user account.
Then step S202, i.e. acquisition user account data include step：
Obtain the user account data of doubtful abnormal user account.
Specifically, obtaining the user account data of doubtful abnormal user account, pre-set user account number classification device can be used
Doubtful abnormal user account is detected, and determines whether doubtful abnormal user account is abnormal user account.
In the above-described embodiments, user account shipping address information is obtained according to user account data；User account is received
Goods address information is segmented, and word segmentation result is obtained, and word segmentation result is input in Clustering Model and obtains classification results, according to point
Class result obtains shipping address similarity；Doubtful abnormal user account is obtained according to shipping address similarity, then using pre-
If detecting the user account data of doubtful abnormal user account when user account classifier, it is possible to reduce the detection of user account
Amount, improves the detection efficiency of user account.
In one embodiment, as shown in fig. 6, step S504, i.e., described that the word segmentation result is input to Clustering Model
In obtain classification results, including step：
User account shipping address information is grouped by S602 according to preset condition, total number packets is calculated, according to grouping
Sum calculates clusters number.
Wherein, clusters number refers to the classification number for being used to classify in Clustering Model.
Specifically, user account shipping address information is grouped according to province, city, county and district etc., i.e., different provinces
Part, different cities, different county and districts are respectively different groups, and the group number after statistical packet obtains total number packets.It is then public according to calculating
Formula N=M*1.1 calculates clusters number.Wherein, M is total number packets, and N is clusters number.
S604, obtains the target word of clusters number as initial cluster center from word segmentation result, using initial cluster center as
Current cluster center.
Specifically, after determining that clusters number is N number of the N number of cluster centre of random initializtion is obtained from word segmentation result
N number of target word is used as initial cluster center, using the initial cluster center as current cluster center.
S606 obtains other words in addition to target word in word segmentation result, calculates other words in addition to target word to working as
The distance at prevariety center.
Specifically, using Euclidean distance meter from the other words obtained in word segmentation result in addition to target word in word segmentation result
Distance of other words of the calculation in addition to target word to current cluster center.
Other words in addition to target word are assigned in the corresponding cluster in current cluster center according to distance, are gathered by S608
The target cluster of class number.
Specifically, according to other words in addition to target word to the distance at all current cluster centers, it will be in addition to target word
Other words be assigned in the smallest cluster, just obtained the target cluster of clusters number.An other words are obtained to all
The distance at current cluster center, the distance for determining that current cluster center is most short, and just the word is assigned in the current cluster.
S610 calculates the target cluster center of target cluster, using target cluster center as current cluster center, returns to calculate and removes target
Other words other than word to current cluster center apart from the step of carry out repeating cluster being divided when meeting the condition of convergence
Class result.
Specifically, recalculating the target cluster center of target cluster, using target cluster center as current cluster center, returns and calculate
Other words in addition to target word to current cluster center apart from the step of carry out repeating cluster, when meeting the condition of convergence,
When i.e. current cluster center is consistent with last cluster center, that is, meet the condition of convergence, i.e., using target cluster as classification results.Its
In, it can be used SEE (error sum of squares, Sum of Squared Error, abbreviation SSE), each sample point and its affiliated matter
The quadratic sum of the distance of the heart, the objective function as measurement clustering result quality is as cost function, when cost function reaches minimum value
When, current cluster center is consistent with last cluster center.
In one embodiment, the target word of clusters number can be reselected as initial cluster center, carry out cluster meter
It calculates, obtains classification results, compare cost function value, using the smallest classification results of cost function value as Clustering Model.
In above-described embodiment, by the way that user account shipping address information to be grouped according to preset condition, grouping is calculated
Sum calculates clusters number according to total number packets；The target word of clusters number is obtained from word segmentation result as initial cluster center,
Using initial cluster center as current cluster center；Obtain other words in addition to target word in word segmentation result, calculate except target word with
Distance of the outer other words to current cluster center；Other words in addition to target word are assigned to current cluster center pair according to distance
In the cluster answered, the target cluster of clusters number is obtained；The target cluster center for calculating target cluster, using target cluster center as in current cluster
The heart, return calculate other words in addition to target word to current cluster center apart from the step of carry out repeating cluster, until satisfaction
When the condition of convergence, classification results are obtained, make it possible to more convenient and are accurately obtained classification results.
In one embodiment, as shown in fig. 7, user account detection method, further includes step：
S702, obtains input feature value according to user account data, and input feature value is input to pre-set user account
In number detection model, output feature vector is obtained.
Wherein, pre-set user account detection model is to be trained in advance using sample data using logistic regression algorithm
It arrives, for carrying out user account detection.Input feature value includes account primary attribute vector, facility information vector, user
Behavior vector sum business information vector, output feature vector includes testing result vector.
Specifically, carrying out feature extraction according to historical user's account data, input feature vector is obtained, wherein input feature vector packet
Account primary attribute feature, facility information feature, user behavior characteristics and business information feature are included, account primary attribute feature is used
In the essential information for describing the user, such as account name feature, sex character, age characteristics, address feature and cell-phone number feature
Deng.Facility information feature is used to describe the device parameter information of login user account, such as device operating system version number feature,
Device-fingerprint feature, at one's side chip features, hardware characteristics and equipment are in and escape from prison or crack pattern feature etc..User behavior is special
Take over for use makes user account carry out the data generated when various operations in webpage or client in description user.For example, user page
Face stay time feature, the access order feature of user, operating frequency feature and key information feature etc..Business information feature is used
The information characteristics of generation when describing platform and how to carry out business activity.For example, when there is discount coupon activity, then business information feature
It can be coupon information feature, preferential rule feature etc..The corresponding user of input feature vector is being obtained according to user account data
Account data obtains input feature value according to the corresponding user account data of input feature vector, the input feature value that will be obtained
It is input in pre-set user account detection model, obtains output feature vector.
S704 obtains user account testing result according to output feature vector.
Specifically, the corresponding output feature of various testing results is obtained previously according to testing result, then according to data spy
Obtain corresponding output feature vector, when detecting, according to the output feature vector that detection obtains obtain output feature to
Corresponding output feature is measured, corresponding testing result is obtained according to output feature.For example, being previously obtained user account testing result
Corresponding output feature is 1 when for normal users account, corresponding defeated when user account testing result is abnormal user account
Feature is 0 out, and the corresponding output feature vector of obtained normal users account is [1], the corresponding abnormal use of abnormal user account
Family account [0].
In the above-described embodiments, by obtaining input feature value according to user account data, input feature value is defeated
Enter into pre-set user account detection model, obtains output feature vector；User account detection is obtained according to output feature vector
As a result, can effectively detect abnormal user account using pre-set user account detection model.
In one embodiment, as shown in figure 8, the generation step of the pre-set user account detection model, including step：
S802 obtains historical user's account data and corresponding testing result, is gone through according to historical user's account data
History input feature value obtains history output feature vector according to testing result.
Specifically, according to the corresponding input feature vector number of input feature vector that historical user's account data is extracted in advance
According to, input feature value is obtained according to input feature vector data, obtains corresponding output feature according to testing result, it is special according to output
Obtain history output feature vector.
S804, using history input feature value as the input of Logic Regression Models, using history output feature vector as
The label of Logic Regression Models is trained.
Specifically, using history input feature value as the input of Logic Regression Models, the Logic Regression Models be using
What Sigmoid function was established, wherein Sigmoid function isIt is returned using history output feature vector as logic
The label of model is returned to be trained.
S806 obtains pre-set user account detection model when the cost function of Logic Regression Models reaches preset threshold.
Wherein, cost function is the function for measuring the difference between the value and true value that model prediction comes out.
Specifically, cross entropy can be used as cost function in the cost function of Logic Regression Models, wherein cross entropy letter
Number isWherein C is difference value, and y is desired output, and a is reality output.Work as C
When reaching preset threshold, illustrates to train completion, just obtained pre-set user account detection model at this time.
In the above-described embodiments, by obtaining historical user's account data and corresponding testing result, according to historical user
Account data obtains history input feature value, history output feature vector is obtained according to testing result, by history input feature vector
Input of the vector as Logic Regression Models is trained using history output feature vector as the label of Logic Regression Models,
When the cost function of Logic Regression Models reaches preset threshold, pre-set user account detection model is obtained, can be trained in advance
Good user account detection model can be carried out directly when detecting using the efficiency for improving user account detection.
It in a specific embodiment, can be to the user account of user's registration when user is when registering user account
It is detected, prevents user's batch registration user account.Specifically, obtaining user account data, including user account basis letter
Breath and facility information, obtain user characteristics attribute according to user account data, specifically, user characteristics attribute includes user mobile phone
Number, device chip information, device-fingerprint information, hardware information, equipment mode information and device operating system model etc., will use
Family characteristic attribute is input in pre-set user account number classification device, obtains output feature, if the corresponding testing result of output feature is
Abnormal user account then illustrates the user account registered as the user account of user's batch registration, at this point it is possible to send out to user
The information for sending registration failure prevents client's batch registration user account.
It should be understood that although each step in the flow chart of Fig. 2-8 is successively shown according to the instruction of arrow,
These steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly stating otherwise herein, these steps
Execution there is no stringent sequences to limit, these steps can execute in other order.Moreover, at least one in Fig. 2-8
Part steps may include that perhaps these sub-steps of multiple stages or stage are not necessarily in synchronization to multiple sub-steps
Completion is executed, but can be executed at different times, the execution sequence in these sub-steps or stage is also not necessarily successively
It carries out, but can be at least part of the sub-step or stage of other steps or other steps in turn or alternately
It executes.
In one embodiment, as shown in figure 9, providing a kind of user account detection device 900, including：Characteristic attribute
Obtain module 902, output feature obtains module 904 and testing result obtains module 906, wherein：
Characteristic attribute obtains module 902, for obtaining user account data, obtains user characteristics according to user account data
Attribute；
Output feature obtains module 904, for user characteristics attribute to be input in pre-set user account number classification device, obtains
Export feature；
Testing result obtains module 906, for obtaining user account testing result according to output feature.
In the above-described embodiments, it obtains obtaining user characteristics according to user account data in module 902 by characteristic attribute
Then attribute obtains that user characteristics attribute being input in pre-set user account number classification device in module 904 and obtains in output feature
Feature is exported, finally obtains obtaining user account testing result according to output feature in module 906 in testing result, it can be effective
The abnormal user account out of detection.
In one embodiment, user account detection device 900 further includes：
Historical data obtains module, for obtaining historical user's account data and corresponding testing result, testing result packet
Include history normal users account and history abnormal user account；
Frequency computing module, for according to historical user's account data and corresponding testing result statistical history user account
Number, history normal users account number and history abnormal user account number, and calculate history normal users account frequency and history is different
Normal user account frequency；
Division module, for obtaining corresponding historical user's feature according to historical user's account data, and to historical user
Feature is divided according to preset condition, obtains item to be sorted；
Conditional probability computing module, for counting the corresponding historical user's account number of item to be sorted, history normal users account
Number and history abnormal user account number, and calculating the corresponding historical user's account of item to be sorted is history normal users account
Conditional probability and historical user's account are the conditional probability of history abnormal user account, obtain pre-set user account number classification device.
In one embodiment, output feature obtains module 904, including：
Target Acquisition module, for obtaining the corresponding target of user characteristics attribute item to be sorted；
Bayes's computing module uses shellfish according to conditional probability for obtaining the corresponding conditional probability of target item to be sorted
This theorem of leaf calculates separately that user account is normal users account probability and user account is abnormal user account probability；
Comparison module is obtained for comparing normal users account probability and abnormal user account probability according to comparison result
Export feature.
In one embodiment, user account detection device 900 further includes：
Address obtains module, for obtaining user account shipping address information according to user account data；
Categorization module obtains word segmentation result for segmenting to user account shipping address information, and word segmentation result is defeated
Enter and obtain classification results into Clustering Model, obtains shipping address similarity according to classification results；
Doubtful account obtains module, for obtaining doubtful abnormal user account according to shipping address similarity；
Characteristic attribute obtains module 902, including：
Doubtful abnormal data obtains module, for obtaining the user account data of doubtful abnormal user account.
In one embodiment, categorization module, including：
Cluster numbers computing module calculates and divides for user account shipping address information to be grouped according to preset condition
Group sum calculates clusters number according to total number packets；
Cluster center determining module, the target word for the acquisition clusters number from word segmentation result, will as initial cluster center
Initial cluster center is as current cluster center；
Distance calculation module calculates in addition to target word for obtaining other words in word segmentation result in addition to target word
Other words to current cluster center distance；
Target cluster obtains module, corresponding for other words in addition to target word to be assigned to current cluster center according to distance
Cluster in, obtain the target cluster of clusters number；
Classification results obtain module, for calculating the target cluster center of target cluster, using target cluster center as in current cluster
The heart, return calculate other words in addition to target word to current cluster center apart from the step of carry out repeating cluster, until satisfaction
When the condition of convergence, classification results are obtained.
In one embodiment, user account detection device 900 further includes：
User account detection module, for obtaining input feature value according to user account data, by input feature value
It is input in pre-set user account detection model, obtains output feature vector；
Testing result obtains module, for obtaining user account testing result according to output feature vector.
In one embodiment, user account detection device 900 further includes：
History vectors obtain module, for obtaining historical user's account data and corresponding testing result, are used according to history
Family account data obtains history input feature value, obtains history output feature vector according to testing result；
Training module, for using history input feature value as the input of Logic Regression Models, history to be exported feature
Vector is trained as the label of Logic Regression Models；
Detection model obtains module, for being preset when the cost function of Logic Regression Models reaches preset threshold
User account detection model.
Specific about user account detection device limits the limit that may refer to above for user account detection method
Fixed, details are not described herein.Modules in above-mentioned user account detection device can fully or partially through software, hardware and its
Combination is to realize.Above-mentioned each module can be embedded in the form of hardware or independently of in the processor in computer equipment, can also be with
It is stored in the memory in computer equipment in a software form, in order to which processor calls the above modules of execution corresponding
Operation.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 10.The computer equipment include by system bus connect processor, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program and data
Library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The database of machine equipment is for storing user account data.The network interface of the computer equipment is used to pass through with external terminal
Network connection communication.To realize a kind of user account detection method when the computer program is executed by processor.
It will be understood by those skilled in the art that structure shown in Figure 10, only part relevant to application scheme
The block diagram of structure, does not constitute the restriction for the computer equipment being applied thereon to application scheme, and specific computer is set
Standby may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment is provided, including memory, processor and storage are on a memory
And the computer program that can be run on a processor, processor realize following steps when executing computer program：Obtain user's account
Number obtains user characteristics attribute according to user account data；User characteristics attribute is input to pre-set user account number classification
In device, output feature is obtained；User account testing result is obtained according to output feature.
In one embodiment, following steps are also realized when processor executes computer program：Obtain historical user's account
Data and corresponding testing result, testing result include history normal users account and history abnormal user account；According to history
User account data and corresponding testing result statistical history user account number, history normal users account number and history are used extremely
Family account number, and calculate history normal users account frequency and history abnormal user account frequency；According to historical user's account number
According to obtaining corresponding historical user's feature, and historical user's feature is divided according to preset condition, obtains item to be sorted；System
The corresponding historical user's account number of item to be sorted, history normal users account number and history abnormal user account number are counted, and is calculated
The corresponding historical user's account of item to be sorted is the conditional probability of history normal users account and historical user's account is that history is different
The conditional probability of normal user account, obtains pre-set user account number classification device.
In one embodiment, following steps are also realized when processor executes computer program：Obtain user characteristics attribute
Corresponding target item to be sorted；The corresponding conditional probability of target item to be sorted is obtained, Bayes' theorem is used according to conditional probability
Calculate separately that user account is normal users account probability and user account is abnormal user account probability；Compare normal users account
Number probability and abnormal user account probability obtain output feature according to comparison result.
In one embodiment, following steps are also realized when processor executes computer program：According to user account data
Obtain user account shipping address information；User account shipping address information is segmented, word segmentation result is obtained, participle is tied
Fruit, which is input in Clustering Model, obtains classification results, obtains shipping address similarity according to classification results；According to shipping address phase
Doubtful abnormal user account is obtained like degree；Following steps are also realized when then processor executes computer program：Obtain doubtful exception
The user account data of user account.
In one embodiment, following steps are also realized when processor executes computer program：By user account place of acceptance
Location information is grouped according to preset condition, calculates total number packets, calculates clusters number according to total number packets；From word segmentation result
The target word of clusters number is obtained as initial cluster center, using initial cluster center as current cluster center；It obtains in word segmentation result
Other words in addition to target word, the distance of other words of the calculating in addition to target word to current cluster center；It will be removed according to distance
Other words other than target word are assigned in the corresponding cluster in current cluster center, obtain the target cluster of clusters number；Calculate target cluster
Target cluster center return using target cluster center as current cluster center and calculate other words in addition to target word to current cluster
Center apart from the step of carry out repeat cluster obtain classification results when meeting the condition of convergence.
In one embodiment, following steps are also realized when processor executes computer program：According to user account data
Input feature value is obtained, input feature value is input in pre-set user account detection model, obtains output feature vector；
User account testing result is obtained according to output feature vector.
In one embodiment, following steps are also realized when processor executes computer program：Obtain historical user's account
Data and corresponding testing result obtain history input feature value according to historical user's account data, according to testing result
Feature vector is exported to history；Using history input feature value as the input of Logic Regression Models, by history export feature to
It measures and is trained as the label of Logic Regression Models；When the cost function of Logic Regression Models reaches preset threshold, obtain
Pre-set user account detection model.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program realizes following steps when being executed by processor：User account data are obtained, obtain user spy according to user account data
Levy attribute；User characteristics attribute is input in pre-set user account number classification device, output feature is obtained；It is obtained according to output feature
User account testing result.
In one embodiment, following steps are also realized when computer program is executed by processor：Obtain historical user's account
Number and corresponding testing result, testing result include history normal users account and history abnormal user account；According to going through
History user account data and corresponding testing result statistical history user account number, history normal users account number and history are abnormal
User account number, and calculate history normal users account frequency and history abnormal user account frequency；According to historical user's account
Data obtain corresponding historical user's feature, and divide to historical user's feature according to preset condition, obtain item to be sorted；
The corresponding historical user's account number of item to be sorted, history normal users account number and history abnormal user account number are counted, and is counted
Calculate that the corresponding historical user's account of item to be sorted is the conditional probability of history normal users account and historical user's account is history
The conditional probability of abnormal user account obtains pre-set user account number classification device.
In one embodiment, following steps are also realized when computer program is executed by processor：Obtain user characteristics category
The corresponding target item to be sorted of property；The corresponding conditional probability of target item to be sorted is obtained, it is fixed using Bayes according to conditional probability
Reason calculates separately that user account is normal users account probability and user account is abnormal user account probability；Compare normal users
Account probability and abnormal user account probability obtain output feature according to comparison result.
In one embodiment, following steps are also realized when computer program is executed by processor：According to user account number
According to obtaining user account shipping address information；User account shipping address information is segmented, word segmentation result is obtained, will be segmented
As a result it is input in Clustering Model and obtains classification results, obtain shipping address similarity according to classification results；According to shipping address
Similarity obtains doubtful abnormal user account；Following steps are also realized when then computer program is executed by processor：It obtains doubtful
The user account data of abnormal user account.
In one embodiment, following steps are also realized when computer program is executed by processor：User account is received
Address information is grouped according to preset condition, calculates total number packets, calculates clusters number according to total number packets；From word segmentation result
The middle target word for obtaining clusters number is as initial cluster center, using initial cluster center as current cluster center；Obtain word segmentation result
In other words in addition to target word, calculate other words in addition to target word to current cluster center distance；It will according to distance
Other words in addition to target word are assigned in the corresponding cluster in current cluster center, obtain the target cluster of clusters number；Calculate target
The target cluster center of cluster returns to other words of the calculating in addition to target word to currently using target cluster center as current cluster center
Cluster center apart from the step of carry out repeat cluster obtain classification results when meeting the condition of convergence.
In one embodiment, following steps are also realized when computer program is executed by processor：According to user account number
According to input feature value is obtained, input feature value is input in pre-set user account detection model, obtain output feature to
Amount；User account testing result is obtained according to output feature vector.
In one embodiment, following steps are also realized when computer program is executed by processor：Obtain historical user's account
Number and corresponding testing result obtain history input feature value according to historical user's account data, according to testing result
Obtain history output feature vector；Using history input feature value as the input of Logic Regression Models, history is exported into feature
Vector is trained as the label of Logic Regression Models；When the cost function of Logic Regression Models reaches preset threshold, obtain
To pre-set user account detection model.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The several embodiments of the application above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the concept of this application, various modifications and improvements can be made, these belong to the protection of the application
Range.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of user account detection method, which is characterized in that the method includes：
User account data are obtained, obtain user characteristics attribute according to the user account data；
The user characteristics attribute is input in pre-set user account number classification device, output feature is obtained；
User account testing result is obtained according to the output feature.
2. the method according to claim 1, wherein the generation step packet of the pre-set user account number classification device
It includes：
Historical user's account data and corresponding testing result are obtained, the testing result includes history normal users account and goes through
History abnormal user account；
According to historical user's account data and corresponding testing result statistical history user account number, history normal users account
Number and history abnormal user account number, and calculate history normal users account frequency and history abnormal user account frequency；
Corresponding historical user's feature is obtained according to historical user's account data, and to historical user's feature according to pre-
If condition is divided, item to be sorted is obtained；
Count the corresponding historical user's account number of the item to be sorted, history normal users account number and history abnormal user account
Number, and calculate conditional probability and historical user that the corresponding historical user's account of the item to be sorted is history normal users account
Account is the conditional probability of history abnormal user account, obtains pre-set user account number classification device.
3. the method according to claim 1, wherein described be input to pre-set user for the user characteristics attribute
In account number classification device, output feature is obtained, including：
Obtain the corresponding target item to be sorted of the user characteristics attribute；
The corresponding conditional probability of target item to be sorted is obtained, is calculated separately according to the conditional probability using Bayes' theorem
User account is normal users account probability and user account is abnormal user account probability；
Compare the normal users account probability and the abnormal user account probability, output feature is obtained according to comparison result.
4. the method according to claim 1, wherein the method, further includes：
User account shipping address information is obtained according to the user account data；
The user account shipping address information is segmented, word segmentation result is obtained, the word segmentation result is input to cluster
Classification results are obtained in model, obtain shipping address similarity according to classification results；
Doubtful abnormal user account is obtained according to the shipping address similarity；
The then acquisition user account data, including：
Obtain the user account data of the doubtful abnormal user account.
5. according to the method described in claim 4, it is characterized in that, the described word segmentation result is input in Clustering Model obtains
To classification results, including：
The user account shipping address information is grouped according to preset condition, total number packets are calculated, according to the grouping
Sum calculates clusters number；
The target word of clusters number is obtained from the word segmentation result as initial cluster center, using the initial cluster center as working as
Prevariety center；
Other words in the word segmentation result in addition to the target word are obtained, are calculated described other in addition to the target word
Distance of the word to the current cluster center；
Other words in addition to target word are assigned in the corresponding cluster in the current cluster center according to the distance, are obtained
The target cluster of the clusters number；
The target cluster center for calculating the target cluster returns and removes described in calculating using target cluster center as current cluster center
Other words other than target word to the current cluster center apart from the step of carry out repeating cluster, until meeting the condition of convergence
When, obtain classification results.
6. the method according to claim 1, wherein the method, further includes：
Input feature value is obtained according to the user account data, the input feature value is input to pre-set user account
In detection model, output feature vector is obtained；
User account testing result is obtained according to the output feature vector.
7. according to the method described in claim 6, it is characterized in that, the generation step of the pre-set user account detection model,
Including：
Historical user's account data and corresponding testing result are obtained, history input is obtained according to historical user's account data
Feature vector obtains history output feature vector according to the testing result；
Using the history input feature value as the input of Logic Regression Models, using history output feature vector as institute
The label for stating Logic Regression Models is trained；
When the cost function of the Logic Regression Models reaches preset threshold, the pre-set user account detection model is obtained.
8. a kind of user account detection device, which is characterized in that described device includes：
Characteristic attribute obtains module, for obtaining user account data, obtains user characteristics category according to the user account data
Property；
Output feature obtains module, for the user characteristics attribute to be input in pre-set user account number classification device, obtains defeated
Feature out；
Testing result obtains module, for obtaining user account testing result according to the output feature.
9. a kind of computer equipment, including memory, processor and it is stored on the memory and can be on the processor
The computer program of operation, which is characterized in that the processor is realized in claim 1 to 7 when executing the computer program
The step of any one the method.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 7 is realized when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810547162.5A CN108898418B (en) | 2018-05-31 | 2018-05-31 | User account detection method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810547162.5A CN108898418B (en) | 2018-05-31 | 2018-05-31 | User account detection method, device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108898418A true CN108898418A (en) | 2018-11-27 |
| CN108898418B CN108898418B (en) | 2023-06-23 |
Family
ID=64343899
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810547162.5A Active CN108898418B (en) | 2018-05-31 | 2018-05-31 | User account detection method, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108898418B (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109784811A (en) * | 2019-01-15 | 2019-05-21 | 长春市震撼科技有限公司 | A kind of commodity sale system for e-commerce |
| CN110018957A (en) * | 2019-02-14 | 2019-07-16 | 阿里巴巴集团控股有限公司 | A kind of money damage verification script detection method and device |
| CN110225036A (en) * | 2019-06-12 | 2019-09-10 | 北京奇艺世纪科技有限公司 | A kind of account detection method, device, server and storage medium |
| CN110427999A (en) * | 2019-07-26 | 2019-11-08 | 武汉斗鱼网络科技有限公司 | A kind of account relevance evaluation method, apparatus, equipment and medium |
| CN110474871A (en) * | 2019-07-05 | 2019-11-19 | 中国平安财产保险股份有限公司 | A kind of exception account detection method, device, computer equipment and storage medium |
| CN111079029A (en) * | 2019-12-20 | 2020-04-28 | 珠海格力电器股份有限公司 | Sensitive account detection method, storage medium and computer equipment |
| CN111242398A (en) * | 2018-11-29 | 2020-06-05 | 北京搜狗科技发展有限公司 | Data processing method and device for data processing |
| WO2020181854A1 (en) * | 2019-03-13 | 2020-09-17 | 北京三快在线科技有限公司 | Payment anomaly detection |
| CN111831825A (en) * | 2020-07-23 | 2020-10-27 | 咪咕文化科技有限公司 | Account detection method, account detection device, network equipment and storage medium |
| CN111985703A (en) * | 2020-08-12 | 2020-11-24 | 支付宝(杭州)信息技术有限公司 | User identity state prediction method, device and equipment |
| WO2020253381A1 (en) * | 2019-06-17 | 2020-12-24 | 深圳壹账通智能科技有限公司 | Data monitoring method and apparatus, computer device and storage medium |
| CN112153426A (en) * | 2020-09-21 | 2020-12-29 | 腾讯科技（深圳）有限公司 | Content account management method and device, computer equipment and storage medium |
| CN112329811A (en) * | 2020-09-18 | 2021-02-05 | 广州三七网络科技有限公司 | Abnormal account identification method and device, computer equipment and storage medium |
| CN113129018A (en) * | 2021-05-17 | 2021-07-16 | 无锡航吴科技有限公司 | Financing platform account classification method and system |
| CN113521750A (en) * | 2021-07-15 | 2021-10-22 | 珠海金山网络游戏科技有限公司 | Abnormal account detection model training method and abnormal account detection method |
| CN113971038A (en) * | 2020-07-22 | 2022-01-25 | 北京达佳互联信息技术有限公司 | Application program account abnormity identification method, device, server and storage medium |
| CN114861177A (en) * | 2022-04-19 | 2022-08-05 | 中国科学院信息工程研究所 | Method and device for detecting suspicious account on social network |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103577991A (en) * | 2012-08-03 | 2014-02-12 | 阿里巴巴集团控股有限公司 | User identification method and device |
| CN104008332A (en) * | 2014-04-30 | 2014-08-27 | 浪潮电子信息产业股份有限公司 | Intrusion detection system based on Android platform |
| CN105787025A (en) * | 2016-02-24 | 2016-07-20 | 腾讯科技（深圳）有限公司 | Network platform public account classifying method and device |
| CN106886518A (en) * | 2015-12-15 | 2017-06-23 | 国家计算机网络与信息安全管理中心 | A kind of method of microblog account classification |
| CN107872436A (en) * | 2016-09-27 | 2018-04-03 | 阿里巴巴集团控股有限公司 | A kind of account recognition methods, apparatus and system |
-
2018
- 2018-05-31 CN CN201810547162.5A patent/CN108898418B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103577991A (en) * | 2012-08-03 | 2014-02-12 | 阿里巴巴集团控股有限公司 | User identification method and device |
| CN104008332A (en) * | 2014-04-30 | 2014-08-27 | 浪潮电子信息产业股份有限公司 | Intrusion detection system based on Android platform |
| CN106886518A (en) * | 2015-12-15 | 2017-06-23 | 国家计算机网络与信息安全管理中心 | A kind of method of microblog account classification |
| CN105787025A (en) * | 2016-02-24 | 2016-07-20 | 腾讯科技（深圳）有限公司 | Network platform public account classifying method and device |
| CN107872436A (en) * | 2016-09-27 | 2018-04-03 | 阿里巴巴集团控股有限公司 | A kind of account recognition methods, apparatus and system |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111242398A (en) * | 2018-11-29 | 2020-06-05 | 北京搜狗科技发展有限公司 | Data processing method and device for data processing |
| CN109784811A (en) * | 2019-01-15 | 2019-05-21 | 长春市震撼科技有限公司 | A kind of commodity sale system for e-commerce |
| CN110018957A (en) * | 2019-02-14 | 2019-07-16 | 阿里巴巴集团控股有限公司 | A kind of money damage verification script detection method and device |
| CN110018957B (en) * | 2019-02-14 | 2024-04-09 | 创新先进技术有限公司 | Method and device for detecting resource loss check script |
| WO2020181854A1 (en) * | 2019-03-13 | 2020-09-17 | 北京三快在线科技有限公司 | Payment anomaly detection |
| CN110225036A (en) * | 2019-06-12 | 2019-09-10 | 北京奇艺世纪科技有限公司 | A kind of account detection method, device, server and storage medium |
| WO2020253381A1 (en) * | 2019-06-17 | 2020-12-24 | 深圳壹账通智能科技有限公司 | Data monitoring method and apparatus, computer device and storage medium |
| CN110474871A (en) * | 2019-07-05 | 2019-11-19 | 中国平安财产保险股份有限公司 | A kind of exception account detection method, device, computer equipment and storage medium |
| CN110474871B (en) * | 2019-07-05 | 2023-10-13 | 中国平安财产保险股份有限公司 | Abnormal account detection method and device, computer equipment and storage medium |
| CN110427999A (en) * | 2019-07-26 | 2019-11-08 | 武汉斗鱼网络科技有限公司 | A kind of account relevance evaluation method, apparatus, equipment and medium |
| CN110427999B (en) * | 2019-07-26 | 2022-02-22 | 武汉斗鱼网络科技有限公司 | Account correlation evaluation method, device, equipment and medium |
| CN111079029A (en) * | 2019-12-20 | 2020-04-28 | 珠海格力电器股份有限公司 | Sensitive account detection method, storage medium and computer equipment |
| CN111079029B (en) * | 2019-12-20 | 2023-11-21 | 珠海格力电器股份有限公司 | Sensitive account detection method, storage medium and computer equipment |
| CN113971038A (en) * | 2020-07-22 | 2022-01-25 | 北京达佳互联信息技术有限公司 | Application program account abnormity identification method, device, server and storage medium |
| CN111831825A (en) * | 2020-07-23 | 2020-10-27 | 咪咕文化科技有限公司 | Account detection method, account detection device, network equipment and storage medium |
| CN111831825B (en) * | 2020-07-23 | 2024-03-15 | 咪咕文化科技有限公司 | Account detection method, device, network equipment and storage medium |
| CN111985703A (en) * | 2020-08-12 | 2020-11-24 | 支付宝(杭州)信息技术有限公司 | User identity state prediction method, device and equipment |
| CN111985703B (en) * | 2020-08-12 | 2022-07-29 | 支付宝(杭州)信息技术有限公司 | User identity state prediction method, device and equipment |
| CN112329811A (en) * | 2020-09-18 | 2021-02-05 | 广州三七网络科技有限公司 | Abnormal account identification method and device, computer equipment and storage medium |
| CN112329811B (en) * | 2020-09-18 | 2024-07-26 | 广州三七网络科技有限公司 | Abnormal account identification method, device, computer equipment and storage medium |
| CN112153426B (en) * | 2020-09-21 | 2023-08-29 | 腾讯科技（深圳）有限公司 | Content account management method and device, computer equipment and storage medium |
| CN112153426A (en) * | 2020-09-21 | 2020-12-29 | 腾讯科技（深圳）有限公司 | Content account management method and device, computer equipment and storage medium |
| CN113129018B (en) * | 2021-05-17 | 2023-12-08 | 无锡航吴科技有限公司 | Financing platform account classification method and system |
| CN113129018A (en) * | 2021-05-17 | 2021-07-16 | 无锡航吴科技有限公司 | Financing platform account classification method and system |
| CN113521750B (en) * | 2021-07-15 | 2023-10-24 | 珠海金山数字网络科技有限公司 | Abnormal account detection model training method and abnormal account detection method |
| CN113521750A (en) * | 2021-07-15 | 2021-10-22 | 珠海金山网络游戏科技有限公司 | Abnormal account detection model training method and abnormal account detection method |
| CN114861177A (en) * | 2022-04-19 | 2022-08-05 | 中国科学院信息工程研究所 | Method and device for detecting suspicious account on social network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108898418B (en) | 2023-06-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108898418A (en) | User account detection method, device, computer equipment and storage medium | |
| CN109829776B (en) | Merchant risk assessment method, device, computer equipment and storage medium | |
| CN108769026A (en) | User account detecting system and method | |
| CN111814910B (en) | Abnormality detection method, abnormality detection device, electronic device, and storage medium | |
| WO2020062702A9 (en) | Method and device for sending text messages, computer device and storage medium | |
| CN110570312B (en) | Sample data acquisition method and device, computer equipment and readable storage medium | |
| CN111340086B (en) | Processing method, system, medium and terminal of label-free electronic transaction data | |
| CN113435900A (en) | Transaction risk determination method and device and server | |
| CN114581233A (en) | Money laundering account determining method, apparatus, device, storage medium and program product | |
| CN112818868B (en) | Method and device for identifying illegal user based on behavior sequence characteristic data | |
| CN117151855A (en) | Fraud risk prediction method, apparatus, computer device, and readable storage medium | |
| CN112990989A (en) | Value prediction model input data generation method, device, equipment and medium | |
| CN111340622A (en) | Abnormal transaction cluster detection method and device | |
| CN116402354A (en) | Evaluation parameter determining method and device, medium and electronic equipment | |
| CN114723554B (en) | Abnormal account identification method and device | |
| CN108629506A (en) | Modeling method, device, computer equipment and the storage medium of air control model | |
| US11797961B2 (en) | Vectorization of transactions | |
| CN114820003A (en) | Pricing information abnormity identification method and device, electronic equipment and storage medium | |
| CN108510071B (en) | Data feature extraction method and device and computer readable storage medium | |
| Kang | Fraud Detection in Mobile Money Transactions Using Machine Learning | |
| US12079375B2 (en) | Automatic segmentation using hierarchical timeseries analysis | |
| Goyal et al. | Credit Card Fraud Detection using Logistic Regression and Decision Tree | |
| US20240185090A1 (en) | Assessment of artificial intelligence errors using machine learning | |
| CN115034873A (en) | Loss report type identification method, loss report type identification device, loss report type identification equipment, storage medium and program product | |
| CN114648670A (en) | Weight abnormal waybill identification model construction method and device and computer equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |