CN110083750A - Blacklist screening method, device, computer equipment and storage medium - Google Patents
Blacklist screening method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN110083750A CN110083750A CN201910198221.7A CN201910198221A CN110083750A CN 110083750 A CN110083750 A CN 110083750A CN 201910198221 A CN201910198221 A CN 201910198221A CN 110083750 A CN110083750 A CN 110083750A
- Authority
- CN
- China
- Prior art keywords
- user
- information
- annotated sequence
- promise
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000012216 screening Methods 0.000 title claims abstract description 41
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 49
- 238000012549 training Methods 0.000 claims abstract description 35
- 230000009193 crawling Effects 0.000 claims abstract description 10
- 241000270322 Lepidosauria Species 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 14
- 210000004218 nerve net Anatomy 0.000 claims description 2
- 238000007405 data analysis Methods 0.000 abstract description 4
- 238000004141 dimensional analysis Methods 0.000 abstract description 4
- BQCADISMDOOEFD-UHFFFAOYSA-N Silver Chemical compound [Ag] BQCADISMDOOEFD-UHFFFAOYSA-N 0.000 description 24
- 238000011161 development Methods 0.000 description 24
- 239000003814 drug Substances 0.000 description 24
- 229940079593 drug Drugs 0.000 description 24
- 229910052709 silver Inorganic materials 0.000 description 24
- 239000004332 silver Substances 0.000 description 24
- 238000010586 diagram Methods 0.000 description 16
- 230000006870 function Effects 0.000 description 7
- 230000008859 change Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 239000000203 mixture Substances 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 239000010985 leather Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011002 quantification Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000013475 authorization Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000000284 resting effect Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- 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
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0609—Buyer or seller confidence or verification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
Abstract
The invention discloses blacklist screening method, device, computer equipment and storage mediums.This method comprises: each user information crawled according to preset user list is labeled respectively, obtain with the one-to-one annotated sequence of each user information, to form annotated sequence set;Training convolutional neural networks are treated by annotated sequence set to be trained, and obtain convolutional neural networks model;If detecting target user's information of people's probability to be predicted of breaking one's promise, the corresponding target annotated sequence of target user's information is obtained;Target annotated sequence is input to convolutional neural networks model, target natural person corresponding with target user's information is calculated and breaks one's promise probability;And if target natural person breaks one's promise probability greater than preset people's probability threshold value of breaking one's promise, and is added to blacklist inventory for target user's information is corresponding.This method by crawling user data automatically, using big data analysis multi-dimensional data, realizes multi dimensional analysis user and breaks one's promise risk.
Description
Technical field
The present invention relates to intelligent decision field more particularly to a kind of blacklist screening method, device, computer equipment and deposit
Storage media.
Background technique
Currently, often resting on single aspect when carrying out credit analysis for user, such as the finance of acquisition user exceedes
Phase information, individual delinquency's information etc., the information that can not be broken one's promise with the comprehensive acquisition client of multiple angles carry out comprehensive analysis user
Information Meter, namely credit rating cannot be filtered out from specified user list according to comprehensive user information degree and be lower than default credit
Spend the target user of threshold value.
Summary of the invention
The embodiment of the invention provides a kind of blacklist screening method, device, computer equipment and storage mediums, it is intended to solve
It is certainly directed to user in the prior art and carries out being to filter out credit rating based on single dimension to be lower than default credit rating when credit analysis
The target user of threshold value, can not comprehensively consider various dimensions the problem of.
In a first aspect, the embodiment of the invention provides a kind of blacklist screening methods comprising:
Each user information crawled according to preset user list is labeled respectively, is obtained and each user information one
One corresponding annotated sequence, to form annotated sequence set;Wherein, the user information of each user is at least wrapped in the user list
Include 2 attribute information;
Using each annotated sequence in the annotated sequence set as the input to training convolutional neural networks, by the mark
The corresponding natural person of each annotated sequence breaks one's promise probability as the output to training convolutional neural networks, to institute in note arrangement set
It states and is trained to training convolutional neural networks, obtained for predicting that natural person breaks one's promise the convolutional neural networks model of probability;
If detecting target user's information of people's probability to be predicted of breaking one's promise, the corresponding target of target user's information is obtained
Annotated sequence;
The target annotated sequence is input to the convolutional neural networks model, is calculated and believes with the target user
Corresponding target natural person is ceased to break one's promise probability;
If the target natural person breaks one's promise, probability is greater than preset people's probability threshold value of breaking one's promise, by target user's information pair
That answers is added to blacklist inventory.
Second aspect, the embodiment of the invention provides a kind of blacklist screening apparatus comprising:
Gather acquiring unit, for each user information crawled according to preset user list to be labeled respectively,
Obtain with the one-to-one annotated sequence of each user information, to form annotated sequence set;Wherein, it is respectively used in the user list
The user information at family includes at least 2 attribute information;
Model training unit, for using each annotated sequence in the annotated sequence set as to training convolutional nerve net
The corresponding natural person of annotated sequence each in the annotated sequence set is broken one's promise probability as to training convolutional mind by the input of network
Output through network is trained to described to training convolutional neural networks, is obtained for predicting that natural person breaks one's promise the volume of probability
Product neural network model;
Target sequence acquiring unit, if for detecting target user's information of people's probability to be predicted of breaking one's promise, described in acquisition
The corresponding target annotated sequence of target user's information;
Probability acquiring unit is calculated for the target annotated sequence to be input to the convolutional neural networks model
It breaks one's promise probability to target natural person corresponding with target user's information;And
Inventory updating unit is incited somebody to action if breaking one's promise probability greater than preset people's probability threshold value of breaking one's promise for the target natural person
Target user's information is corresponding to be added to blacklist inventory.
The third aspect, the embodiment of the present invention provide a kind of computer equipment again comprising memory, processor and storage
On the memory and the computer program that can run on the processor, the processor execute the computer program
Blacklist screening method described in the above-mentioned first aspect of Shi Shixian.
Fourth aspect, the embodiment of the invention also provides a kind of computer readable storage mediums, wherein the computer can
It reads storage medium and is stored with computer program, it is above-mentioned that the computer program when being executed by a processor executes the processor
Blacklist screening method described in first aspect.
The embodiment of the invention provides a kind of blacklist screening method, device, computer equipment and storage mediums.This method
Including each user information crawled according to preset user list to be labeled respectively, obtain a pair of with each user information one
The annotated sequence answered, to form annotated sequence set;It is rolled up using each annotated sequence in the annotated sequence set as to training
The input of product neural network, breaks one's promise probability as wait instruct for the corresponding natural person of annotated sequence each in the annotated sequence set
The output for practicing convolutional neural networks, is trained to training convolutional neural networks to described, obtains for predicting that natural person breaks one's promise
The convolutional neural networks model of probability;If detecting target user's information of people's probability to be predicted of breaking one's promise, obtains the target and use
The corresponding target annotated sequence of family information;The target annotated sequence is input to the convolutional neural networks model, is calculated
It breaks one's promise probability to target natural person corresponding with target user's information;And if target natural person probability of breaking one's promise is greater than
Preset people's probability threshold value of breaking one's promise is added to blacklist inventory for target user's information is corresponding.This method passes through automatic
User data is crawled, using big data analysis multi-dimensional data, multi dimensional analysis user is realized and breaks one's promise risk.
Detailed description of the invention
Technical solution in order to illustrate the embodiments of the present invention more clearly, below will be to needed in embodiment description
Attached drawing is briefly described, it should be apparent that, drawings in the following description are some embodiments of the invention, general for this field
For logical technical staff, without creative efforts, it is also possible to obtain other drawings based on these drawings.
Fig. 1 is the application scenarios schematic diagram of blacklist screening method provided in an embodiment of the present invention;
Fig. 2 is the flow diagram of blacklist screening method provided in an embodiment of the present invention;
Fig. 3 is another flow diagram of blacklist screening method provided in an embodiment of the present invention;
Fig. 4 is the sub-process schematic diagram of blacklist screening method provided in an embodiment of the present invention;
Fig. 5 is another sub-process schematic diagram of blacklist screening method provided in an embodiment of the present invention;
Fig. 6 is the schematic block diagram of blacklist screening apparatus provided in an embodiment of the present invention;
Fig. 7 is another schematic block diagram of blacklist screening apparatus provided in an embodiment of the present invention;
Fig. 8 is the subelement schematic block diagram of blacklist screening apparatus provided in an embodiment of the present invention;
Fig. 9 is another subelement schematic block diagram of blacklist screening apparatus provided in an embodiment of the present invention;
Figure 10 is the schematic block diagram of computer equipment provided in an embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are some of the embodiments of the present invention, instead of all the embodiments.Based on this hair
Embodiment in bright, every other implementation obtained by those of ordinary skill in the art without making creative efforts
Example, shall fall within the protection scope of the present invention.
It should be appreciated that ought use in this specification and in the appended claims, term " includes " and "comprising" instruction
Described feature, entirety, step, operation, the presence of element and/or component, but one or more of the other feature, whole is not precluded
Body, step, operation, the presence or addition of element, component and/or its set.
It is also understood that mesh of the term used in this description of the invention merely for the sake of description specific embodiment
And be not intended to limit the present invention.As description of the invention and it is used in the attached claims, unless on
Other situations are hereafter clearly indicated, otherwise " one " of singular, "one" and "the" are intended to include plural form.
It will be further appreciated that the term "and/or" used in description of the invention and the appended claims is
Refer to any combination and all possible combinations of one or more of associated item listed, and including these combinations.
Fig. 1 and Fig. 2 are please referred to, Fig. 1 is the application scenarios schematic diagram of blacklist screening method provided in an embodiment of the present invention,
Fig. 2 is the flow diagram of blacklist screening method provided in an embodiment of the present invention, which is applied to service
In device, this method is executed by the application software being installed in server.
As shown in Fig. 2, the method comprising the steps of S110~S150.
S110, each user information crawled according to preset user list is labeled respectively, is obtained and each user
The one-to-one annotated sequence of information, to form annotated sequence set;Wherein, in the user list each user user information
Including at least 2 attribute information.
In the present embodiment, when user terminal upload user list to server, server is directed to the use that user terminal uploads
Each user crawls user information in name in an account book list, and correspondence obtains annotated sequence after being then labeled to the user information crawled
Set, using annotated sequence set as it is trained be used to predict natural person break one's promise probability convolutional neural networks model data base
Plinth.For the acquisition of various dimensions information relevant to user credit, each user information crawled is believed including at least two attribute
Breath, for example, personal entity's dimension attribute information and business entity's dimension attribute information.
In one embodiment, as shown in figure 3, before step S110 further include:
S101, the corresponding user information of each user in preset user list is crawled by reptile instrument, by what is crawled
User information is stored respectively using the unique identifier of each user as major key to data form.
I.e. after preset user list is uploaded to server by user terminal, server starting reptile instrument crawls preset
The corresponding user information of each user in user list, each user is in addition to including in the user list that wherein user terminal initially uploads
Address name further includes authority corresponding with address name.The user information of two big dimensions is mainly crawled in the process,
First is that personal entity's dimension, second is that business entity's dimension (if i.e. this user also serves as business entity, also needs to analyze its Enterprise Law
The credit rating of people's dimension).
It is individual delinquency for the information that user each in preset user list needs to crawl in personal entity's dimension
Information, non-silver credit information, the overdue information of finance, supreme people's court break one's promise the information such as information;
In business entity's dimension, the required information crawled includes enterprise's actionable information, illegal deception information, State Development and Reform Commission
Information is punished, Bureau of Drugs Supervision's black list information, the information such as great illegal information of tax revenue are eaten;
It is collected by the information to personal entity's dimension and business entity's dimension, can comprehensively analyze the sincerity of user
Degree is single to be all biased from personal entity's dimension or business entity's dimension.
The corresponding user information of each user in preset user list is crawled by reptile instrument, i.e., each user is corresponding
Individual delinquency's information is crawled, non-silver credit information, the overdue information of finance, supreme people's court break one's promise information, enterprise's actionable information, illegal
Deception information, State Development and Reform Commission are punished information, food Bureau of Drugs Supervision's black list information, the illegal information of great tax revenue.It is climbed to store
The above- mentioned information got need to store respectively the user information crawled to number using the unique identifier of each user as major key
According to table.Such as set identification card number for the unique identifier of user, then each user information include identification card number, it is a
People's Crime Information, non-silver credit information, the overdue information of finance, supreme people's court break one's promise information, enterprise's actionable information, illegal deception information,
State Development and Reform Commission is by the information for punishing the fields such as information, food Bureau of Drugs Supervision's black list information, the illegal information of great tax revenue.
In one embodiment, as shown in figure 4, step S101 includes:
S1011, the corresponding authority of each user in the user list is obtained, each award is obtained by image recognition
Weigh the unique identifier in file;
S1012, the stored user information in preset network address is crawled by the reptile instrument, if detecting preset
There are authentication in network address, will authority corresponding with user each in the user list be uploaded to corresponding network address with
Carry out authentication;
If S1013, the authority crawl user information corresponding with the authority by authentication, with
Obtain user information corresponding with user each in the user list;
S1014, using the corresponding unique identifier of user each in the user list as major key, each user is corresponding
User information form a user data, and be stored in corresponding region in data form.
In the present embodiment, after preset user list is uploaded to server by user terminal, server starts crawler work
Tool crawls the corresponding user information of each user in preset user list, each in the user list that wherein user terminal initially uploads
User further includes authority corresponding with address name in addition to including address name.General authority file is identity card picture,
The unique identifier (such as identification card number) in each authority can be obtained by image recognition at this time.
When according to user list at this moment in order to ensure can (such as querying individual Crime Information needs to log in from preset network address
The network address of public security system, inquiry non-silver credit information and the overdue information of finance need to log in credit investigation system and check loan documentation and overdue
Record, inquiry supreme people's court information of breaking one's promise need to log in network address of court system etc.), and when logging in above-mentioned preset network address, Ke Nengxu
User uploads authority, and the authority that user is uploaded is usually the file of identity card picture or scanned copy format, this
Shi Shangchuan and by verifying after can crawl personal entity's dimension of user and the various information of business entity's dimension.It is crawling
In user list after the above- mentioned information of each user, based on the corresponding unique identifier of user each in the user list
The corresponding user information of each user is formed a user data, and is stored in corresponding region in data form by key.
In one embodiment, as shown in figure 5, step S110 includes:
S111, the corresponding user type of each user is obtained according to the user list;
If S112, the corresponding user type of active user are personal user, strategy is marked according to preset personal user and is obtained
Take the field value of each field information in user information corresponding to active user, with according to each field value form with it is current
The corresponding annotated sequence of user;
If S113, the corresponding user type of active user are enterprise customer, strategy is marked according to preset enterprise customer and is obtained
Take the field value of each field information in user information corresponding to active user, with according to each field value form with it is current
The corresponding annotated sequence of user;
S114, by forming annotated sequence set with the one-to-one annotated sequence of each user information.
In one embodiment, step S112 includes:
The corresponding information bar number of each field information in the corresponding user information of active user is counted, to believe as each field
The field value of breath, and annotated sequence corresponding with active user is formed according to each field value.
In the present embodiment, preset personal user marks strategy are as follows: according to individual delinquency's information, non-silver credit information,
The overdue information of finance, supreme people's court are broken one's promise the corresponding information bar number of information, as the value of above-mentioned each field, and by enterprise's lawsuit
Information, illegal deception information, State Development and Reform Commission are punished information, food Bureau of Drugs Supervision's black list information, the illegal information of great tax revenue respectively
Corresponding information bar number takes 0.
Preset enterprise customer marks strategy are as follows: according to individual delinquency's information, non-silver credit information, the overdue information of finance,
Supreme people's court break one's promise information, enterprise's actionable information, illegal deception information, State Development and Reform Commission punished information, food Bureau of Drugs Supervision's black list information,
The great corresponding information bar number of the illegal information of tax revenue, the value as above-mentioned each field.
It is to have at least that two class users, one kind are personal users in user list, another kind of is enterprise customer.Personal user's table
Show that the user does not serve as legal person in enterprise, enterprise customer indicates that the user serves as legal person in enterprise.Such as personal user
User information crawl after, information is broken one's promise respectively according to individual delinquency's information, non-silver credit information, the overdue information of finance, supreme people's court
Corresponding information bar number, as the value of above-mentioned each field, and enterprise's actionable information, illegal deception information, State Development and Reform Commission by
Punishing information, food Bureau of Drugs Supervision's black list information, the self-corresponding information bar number value of the illegal information of great tax revenue is 0.Enterprise customer's
User information is told after crawling according to break one's promise information, enterprise of individual delinquency's information, non-silver credit information, the overdue information of finance, supreme people's court
Dispute information, illegal deception information, State Development and Reform Commission are punished information, food Bureau of Drugs Supervision's black list information, the illegal information of great tax revenue are each
Self-corresponding information bar number, the value as above-mentioned each field.The difference of enterprise customer and personal user is, enterprise customer's
It is illegal that enterprise's actionable information, illegal deception information, State Development and Reform Commission are punished information, food Bureau of Drugs Supervision's black list information, great tax revenue
Information may have non-zero value.
When forming annotated sequence corresponding with each user, exceed according to individual delinquency's information, non-silver credit information, finance
Break one's promise information, enterprise's actionable information, illegal deception information, State Development and Reform Commission of phase information, supreme people's court is punished information, the food black name of Bureau of Drugs Supervision
Single information, the value of the illegal information of great tax revenue this 9 fields sequentially form a sequence.It has obtained with each user information one by one
Corresponding annotated sequence, by by forming annotated sequence set with the one-to-one annotated sequence of each user information.Pass through above-mentioned turn
Change strategy, user credit index parameter will be considered and carry out quantification treatment, is used convenient for following model training.
S120, using each annotated sequence in the annotated sequence set as the input to training convolutional neural networks, will
The corresponding natural person of each annotated sequence breaks one's promise probability as to the defeated of training convolutional neural networks in the annotated sequence set
Out, it is trained, obtains for predicting that natural person breaks one's promise the convolutional neural networks of probability to training convolutional neural networks to described
Model.
In the present embodiment, when being labeled to user information, from individual delinquency's information, non-silver credit information, finance exceedes
Phase information, supreme people's court break one's promise information, enterprise's actionable information, illegal deception information, and State Development and Reform Commission is punished information, eat the black name of Bureau of Drugs Supervision
Single information, the great illegal information of tax revenue are labeled, such as there are individual delinquency's information then to count Crime Information total number and right
It should be labeled according to Crime Information total number, if individual delinquency's information total number is 10, then be by individual delinquency's information labeling
10, there is no non-silver credit informations to be then labeled as 0, and there are 2 overdue information of finance to be then labeled as 2, and there is no supreme people's courts to break one's promise then
It is labeled as 0, there are 8 actionable informations to be then labeled as 8, and there are 11 illegal deception information to be then labeled as 11, and there are 3 development to change
Leather committee is punished information and is then labeled as 3, and there are 1 Tiao Shi Bureau of Drugs Supervision black list informations to be then labeled as 1, and there are 2 great tax revenues are illegal
Information is then labeled as 2, i.e., this user information be labeled after for [10020811312], the natural person of the user is broken one's promise generally
Rate is labeled as 0.91, when being trained to described to training convolutional neural networks by a large amount of data, obtains prediction natural person
It breaks one's promise the convolutional neural networks model of probability.
If S130, the target user's information for detecting people's probability to be predicted of breaking one's promise, it is corresponding to obtain target user's information
Target annotated sequence.
In the present embodiment, when some or multiple users need to break one's promise people's probabilistic forecasting when, equally need to be by awarding
Power file crawls individual delinquency's information of corresponding user, non-silver credit information, the overdue information of finance, and supreme people's court breaks one's promise information, enterprise
Actionable information, illegal deception information, State Development and Reform Commission are punished information, eat Bureau of Drugs Supervision's black list information, the great illegal information of tax revenue,
It is translated into input of the user information sequence as convolutional neural networks model later, is calculated and believes with the target user
Corresponding target natural person is ceased to break one's promise probability.
S140, the target annotated sequence is input to the convolutional neural networks model, be calculated and the target
The corresponding target natural person of user information breaks one's promise probability.
In this embodiment, when the corresponding target annotated sequence of target user's information for obtaining people's probability to be predicted of breaking one's promise
Afterwards, corresponding target natural person can be calculated according to the convolutional neural networks model trained to break one's promise probability, using as
Judge whether the corresponding target user of target user's information of people's probability to be predicted of breaking one's promise is that maximum probability is broken one's promise the foundation of people.
If S150, the target natural person break one's promise, probability is greater than preset people's probability threshold value of breaking one's promise, by the target user
Information is corresponding to be added to blacklist inventory.
In the present embodiment, when people's probability of breaking one's promise (such as 0.92) corresponding with people's probability user information to be predicted of breaking one's promise
Greater than preset people's probability threshold value of breaking one's promise (the people's probability threshold value that will such as break one's promise is set as 0.8), then it represents that user's maximum probability is
Break one's promise people, people's mark of breaking one's promise should be carried out to it, and blacklist inventory is added.
This method, using big data analysis multi-dimensional data, realizes multi dimensional analysis by crawling user data automatically
User breaks one's promise risk.
The embodiment of the present invention also provides a kind of blacklist screening apparatus, and the blacklist screening apparatus is for executing aforementioned black name
Any embodiment of single screening method.Specifically, referring to Fig. 6, Fig. 6 is blacklist screening apparatus provided in an embodiment of the present invention
Schematic block diagram.The blacklist screening apparatus 100 can be configured in server.
As shown in fig. 6, blacklist screening apparatus 100 includes set acquiring unit 110, model training unit 120, target sequence
Column acquiring unit 130, probability acquiring unit 140, inventory updating unit 150.
Gather acquiring unit 110, for marking each user information crawled according to preset user list respectively
Note, obtain with the one-to-one annotated sequence of each user information, to form annotated sequence set;Wherein, in the user list
The user information of each user includes at least 2 attribute information.
In the present embodiment, when user terminal upload user list to server, server is directed to the use that user terminal uploads
Each user crawls user information in name in an account book list, and correspondence obtains annotated sequence after being then labeled to the user information crawled
Set, using annotated sequence set as it is trained be used to predict natural person break one's promise probability convolutional neural networks model data base
Plinth.For the acquisition of various dimensions information relevant to user credit, each user information crawled is believed including at least two attribute
Breath, for example, personal entity's dimension attribute information and business entity's dimension attribute information.
In one embodiment, as shown in fig. 7, blacklist screening apparatus 100 further include:
Information crawler unit 101, for crawling the corresponding user of each user in preset user list by reptile instrument
Information stores the user information crawled respectively to data form using the unique identifier of each user as major key.
I.e. after preset user list is uploaded to server by user terminal, server starting reptile instrument crawls preset
The corresponding user information of each user in user list, each user is in addition to including in the user list that wherein user terminal initially uploads
Address name further includes authority corresponding with address name.The user information of two big dimensions is mainly crawled in the process,
First is that personal entity's dimension, second is that business entity's dimension (if i.e. this user also serves as business entity, also needs to analyze its Enterprise Law
The credit rating of people's dimension).
It is individual delinquency for the information that user each in preset user list needs to crawl in personal entity's dimension
Information, non-silver credit information, the overdue information of finance, supreme people's court break one's promise the information such as information;
In business entity's dimension, the required information crawled includes enterprise's actionable information, illegal deception information, State Development and Reform Commission
Information is punished, Bureau of Drugs Supervision's black list information, the information such as great illegal information of tax revenue are eaten;
It is collected by the information to personal entity's dimension and business entity's dimension, can comprehensively analyze the sincerity of user
Degree is single to be all biased from personal entity's dimension or business entity's dimension.
The corresponding user information of each user in preset user list is crawled by reptile instrument, i.e., each user is corresponding
Individual delinquency's information is crawled, non-silver credit information, the overdue information of finance, supreme people's court break one's promise information, enterprise's actionable information, illegal
Deception information, State Development and Reform Commission are punished information, food Bureau of Drugs Supervision's black list information, the illegal information of great tax revenue.It is climbed to store
The above- mentioned information got need to store respectively the user information crawled to number using the unique identifier of each user as major key
According to table.Such as set identification card number for the unique identifier of user, then each user information include identification card number, it is a
People's Crime Information, non-silver credit information, the overdue information of finance, supreme people's court break one's promise information, enterprise's actionable information, illegal deception information,
State Development and Reform Commission is by the information for punishing the fields such as information, food Bureau of Drugs Supervision's black list information, the illegal information of great tax revenue.
In one embodiment, as shown in figure 8, information crawler unit 101 includes:
Recognition unit 1011 is identified, for obtaining the corresponding authority of each user in the user list, passes through figure
As identification obtains the unique identifier in each authority;
Identity authenticating unit 1012, for crawling the user stored in preset network address letter by the reptile instrument
Breath, if detecting, there are authentications in preset network address, will authority corresponding with user each in the user list
Corresponding network address is uploaded to carry out authentication;
User information acquiring unit 1013, if being crawled and the authorization text for the authority by authentication
The corresponding user information of part, to obtain user information corresponding with user each in the user list;
Information memory cell 1014, for based on the corresponding unique identifier of user each in the user list
The corresponding user information of each user is formed a user data, and is stored in corresponding region in data form by key.
In the present embodiment, after preset user list is uploaded to server by user terminal, server starts crawler work
Tool crawls the corresponding user information of each user in preset user list, each in the user list that wherein user terminal initially uploads
User further includes authority corresponding with address name in addition to including address name.General authority file is identity card picture,
The unique identifier (such as identification card number) in each authority can be obtained by image recognition at this time.
When according to user list at this moment in order to ensure can (such as querying individual Crime Information needs to log in from preset network address
The network address of public security system, inquiry non-silver credit information and the overdue information of finance need to log in credit investigation system and check loan documentation and overdue
Record, inquiry supreme people's court information of breaking one's promise need to log in network address of court system etc.), and when logging in above-mentioned preset network address, Ke Nengxu
User uploads authority, and the authority that user is uploaded is usually the file of identity card picture or scanned copy format, this
Shi Shangchuan and by verifying after can crawl personal entity's dimension of user and the various information of business entity's dimension.It is crawling
In user list after the above- mentioned information of each user, based on the corresponding unique identifier of user each in the user list
The corresponding user information of each user is formed a user data, and is stored in corresponding region in data form by key.
In one embodiment, as shown in figure 9, set acquiring unit 110 includes:
User type judging unit 111, for obtaining the corresponding user type of each user according to the user list;
First ray acquiring unit 112, if being personal user for the corresponding user type of active user, according to preset
Personal user marks the field value that strategy obtains each field information in user information corresponding to active user, according to each
Field value forms annotated sequence corresponding with active user;
Second retrieval unit 113, if being enterprise customer for the corresponding user type of active user, according to preset
Enterprise customer marks the field value that strategy obtains each field information in user information corresponding to active user, according to each
Field value forms annotated sequence corresponding with active user;
Combined sequence unit 114, for by forming annotated sequence set with the one-to-one annotated sequence of each user information.
In one embodiment, First ray acquiring unit 112 is also used to:
The corresponding information bar number of each field information in the corresponding user information of active user is counted, to believe as each field
The field value of breath, and annotated sequence corresponding with active user is formed according to each field value.
In the present embodiment, preset personal user marks strategy are as follows: according to individual delinquency's information, non-silver credit information,
The overdue information of finance, supreme people's court are broken one's promise the corresponding information bar number of information, as the value of above-mentioned each field, and by enterprise's lawsuit
Information, illegal deception information, State Development and Reform Commission are punished information, food Bureau of Drugs Supervision's black list information, the illegal information of great tax revenue respectively
Corresponding information bar number takes 0.
Preset enterprise customer marks strategy are as follows: according to individual delinquency's information, non-silver credit information, the overdue information of finance,
Supreme people's court break one's promise information, enterprise's actionable information, illegal deception information, State Development and Reform Commission punished information, food Bureau of Drugs Supervision's black list information,
The great corresponding information bar number of the illegal information of tax revenue, the value as above-mentioned each field.
It is to have at least that two class users, one kind are personal users in user list, another kind of is enterprise customer.Personal user's table
Show that the user does not serve as legal person in enterprise, enterprise customer indicates that the user serves as legal person in enterprise.Such as personal user
User information crawl after, information is broken one's promise respectively according to individual delinquency's information, non-silver credit information, the overdue information of finance, supreme people's court
Corresponding information bar number, as the value of above-mentioned each field, and enterprise's actionable information, illegal deception information, State Development and Reform Commission by
Punishing information, food Bureau of Drugs Supervision's black list information, the self-corresponding information bar number value of the illegal information of great tax revenue is 0.Enterprise customer's
User information is told after crawling according to break one's promise information, enterprise of individual delinquency's information, non-silver credit information, the overdue information of finance, supreme people's court
Dispute information, illegal deception information, State Development and Reform Commission are punished information, food Bureau of Drugs Supervision's black list information, the illegal information of great tax revenue are each
Self-corresponding information bar number, the value as above-mentioned each field.The difference of enterprise customer and personal user is, enterprise customer's
It is illegal that enterprise's actionable information, illegal deception information, State Development and Reform Commission are punished information, food Bureau of Drugs Supervision's black list information, great tax revenue
Information may have non-zero value.
When forming annotated sequence corresponding with each user, exceed according to individual delinquency's information, non-silver credit information, finance
Break one's promise information, enterprise's actionable information, illegal deception information, State Development and Reform Commission of phase information, supreme people's court is punished information, the food black name of Bureau of Drugs Supervision
Single information, the value of the illegal information of great tax revenue this 9 fields sequentially form a sequence.It has obtained with each user information one by one
Corresponding annotated sequence, by by forming annotated sequence set with the one-to-one annotated sequence of each user information.Pass through above-mentioned turn
Change strategy, user credit index parameter will be considered and carry out quantification treatment, is used convenient for following model training.
Model training unit 120, for using each annotated sequence in the annotated sequence set as to training convolutional mind
The corresponding natural person of annotated sequence each in the annotated sequence set is broken one's promise probability as to training and rolled up by the input through network
The output of product neural network, is trained to described to training convolutional neural networks, obtains for predicting that natural person breaks one's promise probability
Convolutional neural networks model.
In the present embodiment, when being labeled to user information, from individual delinquency's information, non-silver credit information, finance exceedes
Phase information, supreme people's court break one's promise information, enterprise's actionable information, illegal deception information, and State Development and Reform Commission is punished information, eat the black name of Bureau of Drugs Supervision
Single information, the great illegal information of tax revenue are labeled, such as there are individual delinquency's information then to count Crime Information total number and right
It should be labeled according to Crime Information total number, if individual delinquency's information total number is 10, then be by individual delinquency's information labeling
10, there is no non-silver credit informations to be then labeled as 0, and there are 2 overdue information of finance to be then labeled as 2, and there is no supreme people's courts to break one's promise then
It is labeled as 0, there are 8 actionable informations to be then labeled as 8, and there are 11 illegal deception information to be then labeled as 11, and there are 3 development to change
Leather committee is punished information and is then labeled as 3, and there are 1 Tiao Shi Bureau of Drugs Supervision black list informations to be then labeled as 1, and there are 2 great tax revenues are illegal
Information is then labeled as 2, i.e., this user information be labeled after for [10020811312], the natural person of the user is broken one's promise generally
Rate is labeled as 0.91, when being trained to described to training convolutional neural networks by a large amount of data, obtains prediction natural person
It breaks one's promise the convolutional neural networks model of probability.
Target sequence acquiring unit 130, if obtaining institute for detecting target user's information of people's probability to be predicted of breaking one's promise
State the corresponding target annotated sequence of target user's information.
In the present embodiment, when some or multiple users need to break one's promise people's probabilistic forecasting when, equally need to be by awarding
Power file crawls individual delinquency's information of corresponding user, non-silver credit information, the overdue information of finance, and supreme people's court breaks one's promise information, enterprise
Actionable information, illegal deception information, State Development and Reform Commission are punished information, eat Bureau of Drugs Supervision's black list information, the great illegal information of tax revenue,
It is translated into input of the user information sequence as convolutional neural networks model later, is calculated and believes with the target user
Corresponding target natural person is ceased to break one's promise probability.
Probability acquiring unit 140 is calculated for the target annotated sequence to be input to the convolutional neural networks model
Target natural person corresponding with target user's information is obtained to break one's promise probability.
In this embodiment, when the corresponding target annotated sequence of target user's information for obtaining people's probability to be predicted of breaking one's promise
Afterwards, corresponding target natural person can be calculated according to the convolutional neural networks model trained to break one's promise probability, using as
Judge whether the corresponding target user of target user's information of people's probability to be predicted of breaking one's promise is that maximum probability is broken one's promise the foundation of people.
Inventory updating unit 150, if breaking one's promise probability greater than preset people's probability threshold value of breaking one's promise for the target natural person,
Blacklist inventory is added to by target user's information is corresponding.
In the present embodiment, when people's probability of breaking one's promise (such as 0.92) corresponding with people's probability user information to be predicted of breaking one's promise
Greater than preset people's probability threshold value of breaking one's promise (the people's probability threshold value that will such as break one's promise is set as 0.8), then it represents that user's maximum probability is
Break one's promise people, people's mark of breaking one's promise should be carried out to it, and blacklist inventory is added.
The device, using big data analysis multi-dimensional data, realizes multi dimensional analysis by crawling user data automatically
User breaks one's promise risk.
Above-mentioned blacklist screening apparatus can be implemented as the form of computer program, which can be in such as Figure 10
Shown in run in computer equipment.
Referring to Fig. 10, Figure 10 is the schematic block diagram of computer equipment provided in an embodiment of the present invention.The computer is set
Standby 500 be server, and server can be independent server, is also possible to the server cluster of multiple server compositions.
Refering to fig. 10, which includes processor 502, memory and the net connected by system bus 501
Network interface 505, wherein memory may include non-volatile memory medium 503 and built-in storage 504.
The non-volatile memory medium 503 can storage program area 5031 and computer program 5032.The computer program
5032 are performed, and processor 502 may make to execute blacklist screening method.
The processor 502 supports the operation of entire computer equipment 500 for providing calculating and control ability.
The built-in storage 504 provides environment for the operation of the computer program 5032 in non-volatile memory medium 503, should
When computer program 5032 is executed by processor 502, processor 502 may make to execute blacklist screening method.
The network interface 505 is for carrying out network communication, such as the transmission of offer data information.Those skilled in the art can
To understand, structure shown in Figure 10, only the block diagram of part-structure relevant to the present invention program, is not constituted to this hair
The restriction for the computer equipment 500 that bright scheme is applied thereon, specific computer equipment 500 may include than as shown in the figure
More or fewer components perhaps combine certain components or with different component layouts.
Wherein, the processor 502 is for running computer program 5032 stored in memory, to realize following function
Can: each user information crawled according to preset user list is labeled respectively, is obtained a pair of with each user information one
The annotated sequence answered, to form annotated sequence set;It is rolled up using each annotated sequence in the annotated sequence set as to training
The input of product neural network, breaks one's promise probability as wait instruct for the corresponding natural person of annotated sequence each in the annotated sequence set
The output for practicing convolutional neural networks, is trained to training convolutional neural networks to described, obtains for predicting that natural person breaks one's promise
The convolutional neural networks model of probability;If detecting target user's information of people's probability to be predicted of breaking one's promise, obtains the target and use
The corresponding target annotated sequence of family information;The target annotated sequence is input to the convolutional neural networks model, is calculated
It breaks one's promise probability to target natural person corresponding with target user's information;And if target natural person probability of breaking one's promise is greater than
Preset people's probability threshold value of breaking one's promise is added to blacklist inventory for target user's information is corresponding.
In one embodiment, processor 502 is described by each user crawled according to preset user list letter in execution
Breath is labeled, obtain with the one-to-one annotated sequence of each user, the step of to form annotated sequence set before, also execute
Following operation: the corresponding user information of each user in preset user list, the user that will be crawled are crawled by reptile instrument
Information is stored respectively using the unique identifier of each user as major key to data form.
In one embodiment, processor 502 is respectively used in described crawled in preset user list by reptile instrument of execution
The corresponding user information in family stores the user information crawled respectively to number using the unique identifier of each user as major key
According to table step when, perform the following operations: obtaining the corresponding authority of each user in the user list, pass through image
Identification obtains the unique identifier in each authority;It is crawled by the reptile instrument and to be stored in preset network address
User information, if detecting in preset network address there are authentication, will with each user is corresponding in the user list awards
Power file is uploaded to corresponding network address to carry out authentication;If the authority is crawled and is awarded with described by authentication
The corresponding user information of file is weighed, to obtain user information corresponding with user each in the user list;With the user
The corresponding unique identifier of each user is major key in list, and the corresponding user information of each user is formed a number of users
According to, and it is stored in corresponding region in data form.
In one embodiment, processor 502 is described by each user crawled according to preset user list letter in execution
Breath is labeled respectively, is obtained with the one-to-one annotated sequence of each user information, when step to form annotated sequence set,
It performs the following operations: the corresponding user type of each user is obtained according to the user list;If the corresponding user of active user
Type is personal user, marks strategy according to preset personal user and obtains each word in user information corresponding to active user
The field value of segment information, to form annotated sequence corresponding with active user according to each field value;If active user is corresponding
User type be enterprise customer, strategy is marked according to preset enterprise customer and is obtained corresponding to active user in user information
The field value of each field information, to form annotated sequence corresponding with active user according to each field value;By with each use
Information one-to-one annotated sequence in family forms annotated sequence set.
In one embodiment, processor 502 is described according to the current use of preset personal user mark strategy acquisition in execution
The field value of each field information in user information corresponding to family, it is corresponding with active user to be formed according to each field value
Annotated sequence step when, perform the following operations: each field information is corresponding in the corresponding user information of statistics active user
Information bar number forms mark corresponding with active user using the field value as each field information, and according to each field value
Infuse sequence.
It will be understood by those skilled in the art that the embodiment of computer equipment shown in Figure 10 is not constituted to computer
The restriction of equipment specific composition, in other embodiments, computer equipment may include components more more or fewer than diagram, or
Person combines certain components or different component layouts.For example, in some embodiments, computer equipment can only include depositing
Reservoir and processor, in such embodiments, the structure and function of memory and processor are consistent with embodiment illustrated in fig. 10,
Details are not described herein.
It should be appreciated that in embodiments of the present invention, processor 502 can be central processing unit (Central
Processing Unit, CPU), which can also be other general processors, digital signal processor (Digital
Signal Processor, DSP), specific integrated circuit (Application Specific Integrated Circuit,
ASIC), ready-made programmable gate array (Field-Programmable GateArray, FPGA) or other programmable logic devices
Part, discrete gate or transistor logic, discrete hardware components etc..Wherein, general processor can be microprocessor or
The processor is also possible to any conventional processor etc..
Computer readable storage medium is provided in another embodiment of the invention.The computer readable storage medium can be with
For non-volatile computer readable storage medium.The computer-readable recording medium storage has computer program, wherein calculating
Machine program performed the steps of when being executed by processor each user information that will be crawled according to preset user list respectively into
Rower note, obtain with the one-to-one annotated sequence of each user information, to form annotated sequence set;By the annotated sequence collection
Each annotated sequence is as the input to training convolutional neural networks in conjunction, by each annotated sequence in the annotated sequence set
Corresponding natural person breaks one's promise probability as the output to training convolutional neural networks, carries out to described to training convolutional neural networks
Training is obtained for predicting that natural person breaks one's promise the convolutional neural networks model of probability;If detecting people's probability to be predicted of breaking one's promise
Target user's information obtains the corresponding target annotated sequence of target user's information;The target annotated sequence is input to
The convolutional neural networks model is calculated target natural person corresponding with target user's information and breaks one's promise probability;And
If the target natural person breaks one's promise, probability is greater than preset people's probability threshold value of breaking one's promise, by the corresponding addition of target user's information
To blacklist inventory.
In one embodiment, described to be labeled each user information crawled according to preset user list, it obtains
With the one-to-one annotated sequence of each user, before forming annotated sequence set, further includes: crawled by reptile instrument default
User list in the corresponding user information of each user, be with the unique identifier of each user by the user information crawled
Major key is stored respectively to data form.
In one embodiment, described the corresponding user of each user in preset user list is crawled by reptile instrument to believe
Breath, the user information crawled is stored respectively using the unique identifier of each user as major key to data form, comprising: obtain
The corresponding authority of each user in the user list is taken, unique knowledge in each authority is obtained by image recognition
It does not identify;The stored user information in preset network address is crawled by the reptile instrument, if detecting in preset network address
There are authentications, and authority corresponding with user each in the user list is uploaded to corresponding network address to carry out body
Part verifying;If the authority by authentication, crawls user information corresponding with the authority, to obtain and institute
State the corresponding user information of each user in user list;With the corresponding unique identifier of user each in the user list
For major key, the corresponding user information of each user is formed into a user data, and is stored in corresponding region in data form.
It is in one embodiment, described to be labeled each user information crawled according to preset user list respectively,
Obtain with the one-to-one annotated sequence of each user information, to form annotated sequence set, comprising: obtained according to the user list
Take the corresponding user type of each user;If the corresponding user type of active user is personal user, according to preset personal use
Family mark strategy obtains the field value of each field information in user information corresponding to active user, to be taken according to each field
Value forms annotated sequence corresponding with active user;If the corresponding user type of active user is enterprise customer, according to preset
Enterprise customer marks the field value that strategy obtains each field information in user information corresponding to active user, according to each
Field value forms annotated sequence corresponding with active user;By forming mark with the one-to-one annotated sequence of each user information
Arrangement set.
In one embodiment, described that the letter of user corresponding to active user is obtained according to preset personal user mark strategy
The field value of each field information in breath, to form annotated sequence corresponding with active user according to each field value, comprising:
The corresponding information bar number of each field information in the corresponding user information of active user is counted, using the field as each field information
Value, and annotated sequence corresponding with active user is formed according to each field value.
It is apparent to those skilled in the art that for convenience of description and succinctly, foregoing description is set
The specific work process of standby, device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
Those of ordinary skill in the art may be aware that unit described in conjunction with the examples disclosed in the embodiments of the present disclosure and algorithm
Step can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware and software
Interchangeability generally describes each exemplary composition and step according to function in the above description.These functions are studied carefully
Unexpectedly the specific application and design constraint depending on technical solution are implemented in hardware or software.Professional technician
Each specific application can be used different methods to achieve the described function, but this realization is it is not considered that exceed
The scope of the present invention.
In several embodiments provided by the present invention, it should be understood that disclosed unit and method, it can be with
It realizes by another way.For example, the apparatus embodiments described above are merely exemplary, for example, the unit
It divides, only logical function partition, there may be another division manner in actual implementation, can also will be with the same function
Unit set is at a unit, such as multiple units or components can be combined or can be integrated into another system or some
Feature can be ignored, or not execute.In addition, shown or discussed mutual coupling, direct-coupling or communication connection can
Be through some interfaces, the indirect coupling or communication connection of device or unit, be also possible to electricity, mechanical or other shapes
Formula connection.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.Some or all of unit therein can be selected to realize the embodiment of the present invention according to the actual needs
Purpose.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, is also possible to two or more units and is integrated in one unit.It is above-mentioned integrated
Unit both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in one storage medium.Based on this understanding, technical solution of the present invention is substantially in other words to existing
The all or part of part or the technical solution that technology contributes can be embodied in the form of software products, should
Computer software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be
Personal computer, server or network equipment etc.) execute all or part of step of each embodiment the method for the present invention
Suddenly.And storage medium above-mentioned include: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), magnetic disk or
The various media that can store program code such as person's CD.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can readily occur in various equivalent modifications or replace
It changes, these modifications or substitutions should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with right
It is required that protection scope subject to.
Claims (10)
1. a kind of blacklist screening method characterized by comprising
Each user information crawled according to preset user list is labeled respectively, is obtained a pair of with each user information one
The annotated sequence answered, to form annotated sequence set;Wherein, the user information of each user includes at least 2 in the user list
Attribute information;
Using each annotated sequence in the annotated sequence set as the input to training convolutional neural networks, by the mark sequence
The corresponding natural person of each annotated sequence breaks one's promise probability as the output to training convolutional neural networks in column set, to it is described to
Training convolutional neural networks are trained, and are obtained for predicting that natural person breaks one's promise the convolutional neural networks model of probability;
If detecting target user's information of people's probability to be predicted of breaking one's promise, the corresponding target mark of target user's information is obtained
Sequence;
The target annotated sequence is input to the convolutional neural networks model, is calculated and target user's information pair
The target natural person answered breaks one's promise probability;And
If the target natural person breaks one's promise, probability is greater than preset people's probability threshold value of breaking one's promise, and target user's information is corresponding
It is added to blacklist inventory.
2. blacklist screening method according to claim 1, which is characterized in that it is described will be according to preset user list institute
Each user information crawled is labeled, obtain with the one-to-one annotated sequence of each user, with form annotated sequence set it
Before, further includes:
Crawl the corresponding user information of each user in preset user list by reptile instrument, by the user information crawled with
The unique identifier of each user is that major key is stored respectively to data form.
3. blacklist screening method according to claim 2, which is characterized in that it is described crawled by reptile instrument it is preset
The corresponding user information of each user in user list, by the user information crawled based on the unique identifier of each user
Key is stored respectively to data form, comprising:
The corresponding authority of each user in the user list is obtained, is obtained in each authority by image recognition
Unique identifier;
The stored user information in preset network address is crawled by the reptile instrument, is existed in preset network address if detecting
Authority corresponding with user each in the user list is uploaded to corresponding network address and is tested with carrying out identity by authentication
Card;
If the authority by authentication, crawls user information corresponding with the authority, with obtain with it is described
The corresponding user information of each user in user list;
Using the corresponding unique identifier of user each in the user list as major key, by the corresponding user information of each user
A user data is formed, and is stored in corresponding region in data form.
4. blacklist screening method according to claim 1, which is characterized in that it is described will be according to preset user list institute
Each user information crawled is labeled respectively, obtain with the one-to-one annotated sequence of each user information, to form mark sequence
Column set, comprising:
The corresponding user type of each user is obtained according to the user list;
If the corresponding user type of active user is personal user, strategy is marked according to preset personal user and obtains active user
The field value of each field information in corresponding user information, it is corresponding with active user to be formed according to each field value
Annotated sequence;
If the corresponding user type of active user is enterprise customer, strategy is marked according to preset enterprise customer and obtains active user
The field value of each field information in corresponding user information, it is corresponding with active user to be formed according to each field value
Annotated sequence;
By forming annotated sequence set with the one-to-one annotated sequence of each user information.
5. blacklist screening method according to claim 4, which is characterized in that described to be marked according to preset personal user
Strategy obtains the field value of each field information in user information corresponding to active user, to be formed according to each field value
Annotated sequence corresponding with active user, comprising:
The corresponding information bar number of each field information in the corresponding user information of active user is counted, using as each field information
Field value, and annotated sequence corresponding with active user is formed according to each field value.
6. a kind of blacklist screening apparatus characterized by comprising
Set acquiring unit is obtained for each user information crawled according to preset user list to be labeled respectively
With the one-to-one annotated sequence of each user information, to form annotated sequence set;Wherein, each user in the user list
User information includes at least 2 attribute information;
Model training unit, for using each annotated sequence in the annotated sequence set as to training convolutional neural networks
Input, breaks one's promise probability as to training convolutional nerve net for the corresponding natural person of annotated sequence each in the annotated sequence set
The output of network is trained to described to training convolutional neural networks, obtain for predict natural person break one's promise probability convolution mind
Through network model;
Target sequence acquiring unit, if obtaining the target for detecting target user's information of people's probability to be predicted of breaking one's promise
The corresponding target annotated sequence of user information;
Probability acquiring unit, for the target annotated sequence to be input to the convolutional neural networks model, be calculated with
The corresponding target natural person of target user's information breaks one's promise probability;And
Inventory updating unit will be described if breaking one's promise probability greater than preset people's probability threshold value of breaking one's promise for the target natural person
Target user's information is corresponding to be added to blacklist inventory.
7. blacklist screening apparatus according to claim 6, which is characterized in that further include:
Information crawler unit will for crawling the corresponding user information of each user in preset user list by reptile instrument
The user information crawled is stored respectively using the unique identifier of each user as major key to data form.
8. blacklist screening apparatus according to claim 6, which is characterized in that the set acquiring unit, comprising:
User type judging unit, for obtaining the corresponding user type of each user according to the user list;
First ray acquiring unit, if being personal user for the corresponding user type of active user, according to preset personal use
Family mark strategy obtains the field value of each field information in user information corresponding to active user, to be taken according to each field
Value forms annotated sequence corresponding with active user;
Second retrieval unit is used if being enterprise customer for the corresponding user type of active user according to preset enterprise
Family mark strategy obtains the field value of each field information in user information corresponding to active user, to be taken according to each field
Value forms annotated sequence corresponding with active user;
Combined sequence unit, for by forming annotated sequence set with the one-to-one annotated sequence of each user information.
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 realizes such as claim 1 to 5 when executing the computer program
Any one of described in blacklist screening method.
10. a kind of computer readable storage medium, which is characterized in that the computer-readable recording medium storage has computer journey
Sequence, the computer program execute the processor as described in any one of claim 1 to 5 black
List screening method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910198221.7A CN110083750A (en) | 2019-03-15 | 2019-03-15 | Blacklist screening method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910198221.7A CN110083750A (en) | 2019-03-15 | 2019-03-15 | Blacklist screening method, device, computer equipment and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110083750A true CN110083750A (en) | 2019-08-02 |
Family
ID=67412591
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910198221.7A Pending CN110083750A (en) | 2019-03-15 | 2019-03-15 | Blacklist screening method, device, computer equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110083750A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110597984A (en) * | 2019-08-12 | 2019-12-20 | 大箴(杭州)科技有限公司 | Method and device for determining abnormal behavior user information, storage medium and terminal |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102880934A (en) * | 2012-09-07 | 2013-01-16 | 中国标准化研究院 | Integrity evaluation method for food enterprise |
CN107633081A (en) * | 2017-09-26 | 2018-01-26 | 浙江极赢信息技术有限公司 | A kind of querying method and system of user profile of breaking one's promise |
CN107798068A (en) * | 2017-09-26 | 2018-03-13 | 浙江极赢信息技术有限公司 | A kind of processing method, system and the relevant apparatus of user data of breaking one's promise |
CN109118013A (en) * | 2018-08-29 | 2019-01-01 | 黑龙江工业学院 | A kind of management data prediction technique, readable storage medium storing program for executing and forecasting system neural network based |
CN109272202A (en) * | 2018-08-24 | 2019-01-25 | 中国科学院大学 | A kind of enterprise credit risk method and system based on convolutional neural networks |
-
2019
- 2019-03-15 CN CN201910198221.7A patent/CN110083750A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102880934A (en) * | 2012-09-07 | 2013-01-16 | 中国标准化研究院 | Integrity evaluation method for food enterprise |
CN107633081A (en) * | 2017-09-26 | 2018-01-26 | 浙江极赢信息技术有限公司 | A kind of querying method and system of user profile of breaking one's promise |
CN107798068A (en) * | 2017-09-26 | 2018-03-13 | 浙江极赢信息技术有限公司 | A kind of processing method, system and the relevant apparatus of user data of breaking one's promise |
CN109272202A (en) * | 2018-08-24 | 2019-01-25 | 中国科学院大学 | A kind of enterprise credit risk method and system based on convolutional neural networks |
CN109118013A (en) * | 2018-08-29 | 2019-01-01 | 黑龙江工业学院 | A kind of management data prediction technique, readable storage medium storing program for executing and forecasting system neural network based |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110597984A (en) * | 2019-08-12 | 2019-12-20 | 大箴(杭州)科技有限公司 | Method and device for determining abnormal behavior user information, storage medium and terminal |
CN110597984B (en) * | 2019-08-12 | 2022-05-20 | 大箴(杭州)科技有限公司 | Method and device for determining abnormal behavior user information, storage medium and terminal |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104615852B (en) | The method for order and the raising source service efficiency of registering for guarantee online booking | |
CN104836781B (en) | Distinguish the method and device for accessing user identity | |
CN109783581A (en) | Right management method, device, electronic equipment and storage medium | |
CN106228386A (en) | A kind of information-pushing method and device | |
CN108335214A (en) | Self-service Claims Resolution method, server and computer readable storage medium | |
CN109598628B (en) | Method, device and equipment for identifying medical insurance fraud behaviors and readable storage medium | |
CN109543447A (en) | Information processing method, server and computer storage medium based on block chain | |
CN108269087A (en) | The processing method and processing device of location information | |
CN110245841A (en) | A kind of net about vehicle risk control method and relevant apparatus | |
Shema | Effective credit scoring using limited mobile phone data | |
CN107918911A (en) | System and method for performing safe web bank transaction | |
CN109656960A (en) | Intelligent recommendation method, apparatus, terminal device and the storage medium of credit operation | |
US20090063236A1 (en) | Unidirectionally protected, fully automated asset allocation and asset monitoring apparatuses, and a corresponding method | |
Wein et al. | Analyzing the homeland security of the US‐Mexico border | |
CN113872952A (en) | Identity verification product pushing method, device, equipment and system architecture | |
CN110097170A (en) | Information pushes object prediction model acquisition methods, terminal and storage medium | |
CN110060750A (en) | Medical data method for pushing, system, computer equipment and readable storage medium storing program for executing | |
CN108564366A (en) | Payment cipher remapping method, device and electronic equipment | |
CN110689443A (en) | Insurance data processing method and device, storage medium and server | |
US20150026660A1 (en) | Methods for building application intelligence into event driven applications through usage learning, and systems supporting such applications | |
CN104704521B (en) | Multifactor profile and security fingerprint analysis | |
CN112702410B (en) | Evaluation system, method and related equipment based on blockchain network | |
CN112511632B (en) | Object pushing method, device and equipment based on multi-source data and storage medium | |
CN110083750A (en) | Blacklist screening method, device, computer equipment and storage medium | |
CN109147276A (en) | monitoring 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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190802 |