CN110458425A - Risk analysis method, device, readable medium and the electronic equipment of risk subject - Google Patents
Risk analysis method, device, readable medium and the electronic equipment of risk subject Download PDFInfo
- Publication number
- CN110458425A CN110458425A CN201910678071.XA CN201910678071A CN110458425A CN 110458425 A CN110458425 A CN 110458425A CN 201910678071 A CN201910678071 A CN 201910678071A CN 110458425 A CN110458425 A CN 110458425A
- Authority
- CN
- China
- Prior art keywords
- risk
- subject
- index
- risk index
- information
- 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 76
- 238000012502 risk assessment Methods 0.000 title claims abstract description 73
- 238000012549 training Methods 0.000 claims description 107
- 230000006870 function Effects 0.000 claims description 69
- 238000012360 testing method Methods 0.000 claims description 54
- 238000004422 calculation algorithm Methods 0.000 claims description 43
- 238000005070 sampling Methods 0.000 claims description 39
- 238000004458 analytical method Methods 0.000 claims description 21
- 238000003860 storage Methods 0.000 claims description 17
- 230000000694 effects Effects 0.000 claims description 14
- 239000000284 extract Substances 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000010606 normalization Methods 0.000 claims description 5
- 239000000523 sample Substances 0.000 description 113
- 230000008569 process Effects 0.000 description 21
- 238000010586 diagram Methods 0.000 description 11
- 230000006854 communication Effects 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 238000004519 manufacturing process Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 4
- 208000006011 Stroke Diseases 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000005291 magnetic effect Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000005086 pumping Methods 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 230000029305 taxis Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 206010008190 Cerebrovascular accident Diseases 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 239000012468 concentrated sample Substances 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000033001 locomotion Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 230000007115 recruitment Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the present application is the risk analysis method about a kind of risk subject, device, readable medium and electronic equipment, belongs to Internet technical field, this method comprises: obtaining the relevant information of the risk subject in presumptive area；The relevant information of risk subject is inputted into risk index computation model, obtains the risk index of each risk subject；According to the risk index of each risk subject, the risk subject in the presumptive area is divided into multiple classes, each class therein corresponds to a risk index grade；According to the risk index grade, to carry out risk analysis to the risk subject based on the risk index grade.The technical solution of the embodiment of the present application is by being finely divided risk subject based on risk situation, it is ensured that the careful property and high efficiency of risk analysis.
Description
Technical field
This application involves Internet technical fields, risk analysis method, dress in particular to a kind of risk subject
It sets, readable medium and electronic equipment.
Background technique
Risk subject (such as enterprise) is when carrying out risk analysis, it will usually directly carry out risk point to a large amount of risk
Analysis.Currently, when carrying out the risk analysis of risk subject, although by the way that directly all analysis all risk main body can be certain
Guarantee that risk analysis is comprehensive in degree, but the degree of risk of each risk subject can not be understood and prejudged, leads to wind
It not enough refines, can not be analyzed according to specific aim the case where risk subject, analysis efficiency is low when the analysis of danger.
Summary of the invention
It is a kind of risk analysis method for being designed to provide risk subject of the embodiment of the present application, device, computer-readable
Medium and electronic equipment, and then may insure the careful property and high efficiency of risk analysis at least to a certain extent.
Other characteristics and advantages of the application will be apparent from by the following detailed description, or partially by the application
Practice and acquistion.
According to the one aspect of the embodiment of the present application, a kind of risk analysis method of risk subject is provided, comprising: obtain
The relevant information of risk subject in presumptive area；The relevant information input risk index of the risk subject is calculated into mould
Type obtains the risk index of each risk subject；According to the risk index of each risk subject, the fate will be in
Risk subject in domain is divided into multiple classes, and each class therein corresponds to a risk index grade, to be based on the risk
Index ranking carries out risk analysis to the risk subject.
According to the one aspect of the embodiment of the present application, a kind of risk analysis device of risk subject is provided, comprising: obtain
Module, for obtaining the relevant information for the risk subject being in presumptive area；Prediction module, for by the risk subject
Relevant information inputs risk index computation model, obtains the risk index of each risk subject；Analysis module, for according to
Risk subject in the presumptive area is divided into multiple classes by the risk index of each risk subject, therein each
Class corresponds to a risk index grade, to carry out risk analysis to the risk subject based on the risk index grade.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is used for, according to the risk index
Grade extracts the risk subject of respective numbers respectively from least one class in the multiple class, using as the wind sampled out
Dangerous main body；Risk analysis is carried out based on the risk subject sampled out.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is used for, and obtains each risk
The corresponding sampling probability of index ranking；According to the corresponding sampling probability of each risk index grade, from least
The risk subject of respective numbers is extracted in one class respectively.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is used for, according to each risk
The corresponding sampling probability of index ranking calculates the corresponding sample size of each class；According to each class point
Not corresponding sample size, randomly selects the risk subject of respective numbers respectively from least one class；Or according to described each
The corresponding sample size of class obtains the wind of respective numbers according to the sequence of risk index from high to low from least one class
Dangerous main body.
In some embodiments of the present application, aforementioned schemes are based on, the acquisition module is used for, and acquisition is in described predetermined
The risk analysis data of risk subject in region；By the data normalization for each attribute for including in the risk analysis data
For risk directional information, the risk directional information be used to indicate each attribute data whether Yi Chang information；According to
The risk directional information generates the relevant information of the risk subject.
In some embodiments of the present application, aforementioned schemes are based on, the acquisition module is used for, and is obtained in described predetermined
The relevant information of the risk subject in region within a predetermined period of time.
In some embodiments of the present application, aforementioned schemes are based on, the risk analysis device of the risk subject also wraps
Include: training module, for obtaining the sample training collection of risk subject, each training sample that the sample training is concentrated includes
The relevant information of risk subject and the risk index demarcated for the risk subject；The training sample that the sample training is concentrated
The risk index computation model is trained in this input risk index computation model, so that the risk index meter
The difference between risk index that the risk index and each training sample for calculating each training sample of model output include
Less than predetermined threshold.
In some embodiments of the present application, aforementioned schemes are based on, the risk analysis device of the risk subject also wraps
Include: test module, for obtaining the test sample collection of risk subject, each test sample that the test sample is concentrated includes
The relevant information of risk subject and the risk index demarcated for the risk subject；The test specimens that the test sample is concentrated
The risk index computation model is tested in the risk index computation model after this input training, wherein if instruction
The risk index of each test sample of risk index computation model output after white silk includes with each test sample
Risk index between difference be less than predetermined threshold, it is determined that training after the risk index computation model test passes.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is used for, according to each risk
Risk subject in same risk index section is divided into same by risk index section locating for the risk index of main body
Class, to obtain the multiple class.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is also used to, according to each risk master
The risk index of body generates risk subject list, and each risk subject list corresponds to a risk index grade.
In some embodiments of the present application, aforementioned schemes are based on, the risk index computation model includes that XGBoost is calculated
Method model.
The training module is also used to obtain the sample training collection of risk subject, each instruction that the sample training is concentrated
Practice the relevant information that sample includes risk subject and the risk index for risk subject calibration；
By the objective function of the feature vector input XGBoost algorithm of the training sample, the objective function meter is obtained
The first-loss value of calculation is added to the first regression tree function in the objective function；
According to the first-loss value, the objective function for being followed successively by the XGBoost algorithm adds the second regression tree function,
Until the first regression tree function forecasting risk index calculated and the prediction wind of all second regression tree function calculating
The difference for the risk index that the sum of dangerous index is demarcated with the risk subject is less than predetermined threshold, wherein the forecasting risk refers to
The risk index for the risk subject that the sum of number is predicted for XGBoost algorithm model.
In some embodiments of the present application, aforementioned schemes are based on, the relevant information includes following any one or more
Combination: main body public feelings information, main body operation information, major network platform information and primary influences force information；Wherein, the main body
Public feelings information is used to indicate the public feelings information of risk subject, and the main body operation information is used to indicate the business activities of risk subject
Generated relevant information, the major network platform information are used to indicate the related letter of the associated network platform of risk subject
Breath, the primary influences force information are used to indicate the relevant information of the associated crowd of risk subject.
According to the one aspect of the embodiment of the present application, a kind of computer readable storage medium is provided, is stored thereon with meter
Calculation machine program, which is characterized in that such as above-mentioned risk as described in the examples is realized when the computer program is executed by processor
The risk analysis method of main body.
According to the one aspect of the embodiment of the present application, a kind of electronic equipment is provided, comprising: processor；And memory,
For storing the computer program of the processor；Wherein, the processor is configured to next via the computer program is executed
Execute the risk analysis method such as above-mentioned risk subject as described in the examples.
In the technical solution provided by some embodiments of the present application, fate is calculated by risk index computation model
The risk index of risk subject in domain, and risk subject is divided into multiple classes according to the risk index of each risk subject,
It is then based on each class corresponding risk index grade and risk analysis is carried out to risk subject, it can be according to each risk subject
Risk situation specific aim is monitored, it is ensured that the careful property and high efficiency of risk analysis.
It should be understood that above general description and following detailed description be only it is exemplary and explanatory, not
The application can be limited.
Detailed description of the invention
The drawings herein are incorporated into the specification and forms part of this specification, and shows the implementation for meeting the application
Example, and together with specification it is used to explain the principle of the application.It should be evident that the accompanying drawings in the following description is only the application
Some embodiments for those of ordinary skill in the art without creative efforts, can also basis
These attached drawings obtain other attached drawings.In the accompanying drawings:
Fig. 1 is shown can be using the risk analysis method of the risk subject of the embodiment of the present application or the risk of risk subject
The schematic diagram of the exemplary system architecture of analytical equipment.
Fig. 2 diagrammatically illustrates the process of the risk analysis method of the risk subject of one embodiment according to the application
Figure.
Fig. 3 diagrammatically illustrates the training flow chart of the risk index computation model according to one embodiment of the application.
Fig. 4 diagrammatically illustrates the process of the risk analysis method of the risk subject of one embodiment according to the application
Figure.
Fig. 5 diagrammatically illustrates the process of the risk analysis method of the risk subject of one embodiment according to the application
Figure.
Fig. 6 diagrammatically illustrates the process of the risk analysis method of the risk subject of one embodiment according to the application
Figure.
Fig. 7 shows the signal of the terminal interface of the risk analysis method of the risk subject applied to the embodiment of the present application
Figure.
Fig. 8 diagrammatically illustrates the block diagram of the risk analysis device of the risk subject of one embodiment according to the application.
Fig. 9 shows the structural schematic diagram for being suitable for the computer system for the electronic equipment for being used to realize the embodiment of the present application.
Specific embodiment
Example embodiment is described more fully with reference to the drawings.However, example embodiment can be with a variety of shapes
Formula is implemented, and is not understood as limited to example set forth herein；On the contrary, thesing embodiments are provided so that the application will more
Fully and completely, and by the design of example embodiment comprehensively it is communicated to those skilled in the art.
In addition, described feature, structure or characteristic can be incorporated in one or more implementations in any suitable manner
In example.In the following description, many details are provided to provide and fully understand to embodiments herein.However,
It will be appreciated by persons skilled in the art that the technical solution of the application can be practiced without one or more in specific detail,
Or it can be using other methods, constituent element, device, step etc..In other cases, it is not shown in detail or describes known side
Method, device, realization or operation to avoid fuzzy the application various aspects.
Block diagram shown in the drawings is only functional entity, not necessarily must be corresponding with physically separate entity.
I.e., it is possible to realize these functional entitys using software form, or realized in one or more hardware modules or integrated circuit
These functional entitys, or these functional entitys are realized in heterogeneous networks and/or processor device and/or microcontroller device.
Flow chart shown in the drawings is merely illustrative, it is not necessary to including all content and operation/step,
It is not required to execute by described sequence.For example, some operation/steps can also decompose, and some operation/steps can close
And or part merge, therefore the sequence actually executed is possible to change according to the actual situation.
Fig. 1 is shown can be using the risk analysis method of the risk subject of the embodiment of the present application or the risk of risk subject
The schematic diagram of the exemplary system architecture 100 of analytical equipment.
As shown in Figure 1, system architecture 100 may include terminal device 101, network 102, server 103 and server
104.Network 102 between terminal device 101, server 103 and server 104 to provide the medium of communication link.Network
102 may include various connection types, such as wired, wireless communication link or fiber optic cables etc..
It should be understood that the number of terminal device, network and server in Fig. 1 is only schematical.According to realization need
It wants, can have any number of terminal device, network and server.Such as server 103 and server 104 can be it is multiple
The server cluster etc. of server composition.
User can be used terminal device 101 and be interacted with server 103 and server 104 by network 102, with reception or
Send message etc..Terminal device 101 can be any terminal with computer process ability, including but not limited to: server,
Personal terminal (such as mobile phone, computer).
In the concrete application scene of the application, user can be used terminal device 101 and pass through network 102 to service
Device 103 or server 104 send the request for obtaining the relevant information of the risk subject in presumptive area, with server 103
Or server 104 establishes data acquisition protocols, to ensure to get in presumptive area from server 103 or server 104
Risk subject relevant information.
In one embodiment of the application, terminal device (such as terminal device 101) is being got in presumptive area
Risk subject relevant information after, the relevant information of risk subject can be inputted into risk index computation model, obtained every
The risk index of a risk subject.Then according to the risk index of each risk subject, by the risk in the presumptive area
Main body is divided into multiple classes, and each class therein corresponds to a risk index grade, to be based on the risk index grade pair
The risk subject carries out risk analysis.
In the application one embodiment, terminal device (such as terminal device 101) is defeated by the relevant information of risk subject
Enter risk index meter and calculate model, after obtaining the risk index of each risk subject, according to the risk index of each risk subject
Risk subject list is generated, each risk subject list corresponds to a risk index grade.
In the application one embodiment, terminal device (such as terminal device 101) can to risk index computation model into
Row training, specifically, the sample training collection of the available risk subject of terminal device, each training sample that sample training is concentrated
It include the relevant information of risk subject and the risk index for risk subject calibration；The training sample that sample training is concentrated
Risk index computation model is trained in input risk index computation model, so that the output of risk index computation model is each
The difference between risk index that the risk index of a training sample and each training sample include is less than predetermined threshold.
In the application one embodiment, terminal device (such as terminal device 101) is to risk index computation model training
Afterwards, risk index computation model can also be tested, specifically, the test sample of the available risk subject of terminal device
Collection, each test sample which concentrates include the relevant information of risk subject and the wind for risk subject calibration
Dangerous index；Mould is calculated to risk index in risk index computation model after the test sample input training that test sample is concentrated
Type is tested, wherein if training after risk index computation model output each test sample risk index with it is each
The difference between risk index that test sample includes is less than predetermined threshold, it is determined that the risk index computation model after training is surveyed
Examination is qualified.
It should be noted that the risk analysis method of risk subject provided by the embodiment of the present application can be by terminal device
It executes, correspondingly, the risk analysis device of risk subject is generally positioned in terminal device.But in other realities of the application
It applies in example, server can also have similar function with terminal device, thereby executing risk provided by the embodiment of the present application
The risk analysis scheme of main body.
The realization details of the technical solution of the embodiment of the present application is described in detail below:
Fig. 2 diagrammatically illustrates the process of the risk analysis method of the risk subject of one embodiment according to the application
Figure, the executing subject of the risk analysis method of the risk subject can be the equipment with calculation processing function, such as can be by
Terminal device 101 shown in Fig. 1 executes.As shown in Fig. 2, the risk analysis method of the risk subject includes at least step
S210 is described in detail as follows to step S230:
In step S210, the relevant information of the risk subject in presumptive area is obtained.
In one embodiment of the application, risk subject can be various types of carry out production and operating activities in market
Enterprise, such as company, public institution and school.Presumptive area can refer in same supervision region (such as urban district, block
Deng) or according to the region etc. that supervision demand delimited, do not do particular determination herein.The relevant information of risk subject can be arbitrarily
Information relevant to risk subject, as disclosed information on the internet in risk subject production and operating activities.
The method for obtaining the relevant information of the risk subject in the presumptive area can be through keyword from internet
It is upper to be crawled by keyword, it is also possible to crawl in the database by the cooperation channel pre-established, is not spy herein
It is different to limit.
Since the technical solution of the embodiment of the present application is the wind by the relevant information using risk subject to risk subject
After dangerous index is calculated, after risk subject is classified based on risk index, it is sampled wind respectively from different classes
Dangerous main body, thus in order to guarantee risk subject relevant information obtain convenience, the risk subject relevant information of acquisition it is comprehensive
Property and utility risk subject relevant information calculate the efficiency of analysis, can be by as follows in the embodiment of the present application
Mode obtains the relevant information of risk subject:
In one embodiment of the application, the process of the relevant information of the risk subject in presumptive area is obtained,
May include:
The risk analysis data of risk subject of the acquisition in the presumptive area；
It is risk directional information, the risk in the risk analysis data by the data normalization for each attribute for including
Directional information be used for indicate each attribute data whether Yi Chang information；
The relevant information of the risk subject is generated according to the risk directional information.
In the present embodiment, risk analysis data are the active numbers of institute of the relevant information of risk subject acquired from internet
According to the comprehensive of, it is ensured that relevant information.Risk analysis data include the data of a variety of attributes, wherein a variety of attributes such as net
The negative public sentiment details of network, operation abnormal conditions etc..The data of each attribute of acquisition are marked into sample by crowdsourcing or expert
Mode, which carries out processing, can be normalized to risk directional information, wherein risk directional information, which can be, indicates each attribute
Data whether the information such as label, the character string of Yi Chang predetermined form, data volume can be effectively reduced by normalized
Grade improves the efficiency for calculate using risk subject relevant information analysis on the basis of guaranteeing that information is comprehensive.By risk
Directional information can generate the relevant information of risk subject by way of being sequentially connected in series or being stored as information table.
In one embodiment of the application, the available risk subject in presumptive area is within a predetermined period of time
Relevant information.
In the present embodiment, since predetermined amount of time can be acquiring risk subject relevant information at the time of point to before
The period sometime put is also possible to the period in some time or month.By obtaining the phase in predetermined amount of time
The efficiency of information collection can be guaranteed by closing information.
It should be noted that the technical solution of above-described embodiment can be combined together reality in embodiments herein
It applies.For example the risk analysis data of risk subject within a predetermined period of time in presumptive area can be acquired, then it will adopt
Then the data normalization for each attribute for including in the risk analysis data integrated refers to as risk directional information according to the risk
The relevant information of risk subject is generated to information.
With continued reference to shown in Fig. 2, in step S220, the relevant information input risk index of the risk subject is calculated
Model obtains the risk index of each risk subject.
In one embodiment of the application, risk index computation model is the essence obtained previously according to great amount of samples training
True property reaches the machine learning model of pre-provisioning request, can the relevant information to risk subject carry out calculating analysis output it is corresponding
Risk index.Risk index is gone out according to the relevant information evaluation of the risks such as associated production and operating activities of each risk subject
The problem of risk score, the risk index is higher, risk subject is more, and market management situation is more bad, i.e. the wind of risk subject
Danger is higher.
Since the technical solution of the embodiment of the present application is by the risk using risk index computation model to risk subject
Index is calculated.It also proposed the scheme being trained to risk index computation model in the embodiment of the present application, it is specific such as Fig. 3
It is shown, include the following steps:
Step S310, obtains the sample training collection of risk subject, and each training sample that the sample training is concentrated includes
The relevant information of risky main body and the risk index demarcated for the risk subject；
Step S320, the training sample that the sample training is concentrated input in the risk index computation model to described
Risk index computation model is trained, so that the risk index of each training sample of risk index computation model output
The difference between risk index for including with each training sample is less than predetermined threshold.
In one embodiment of the application, risk index computation model can be set by dictionary table or key-value pair list
All kinds of parameters in the process.Such as, when there are training samples to input in the risk index computation model to the risk index meter
When calculation model is trained, the risk index and the training sample of the training sample of risk index computation model output include
Difference between risk index is greater than predetermined threshold, then can adjust risk index by dictionary table or key-value pair list and calculate mould
The parameters such as the number of the tree of type, so that the risk index and the trained sample of the training sample of risk index computation model output
Originally the difference between risk index for including is less than predetermined threshold.Wherein, the risk index of risk subject calibration can be by expert
It is demarcated, the training effect of model is effectively ensured.
In one embodiment of the application, after being trained to risk index computation model, in order to test training
The effect of risk index computation model afterwards can also test risk index computation model, and detailed process is as follows: obtain
The test sample collection of risk subject, each test sample that the test sample is concentrated include risk subject relevant information and
For the risk index of risk subject calibration；
To described in the risk index computation model after the test sample input training that the test sample is concentrated
Risk index computation model is tested, wherein if each test specimens of the risk index computation model output after training
The difference between risk index that this risk index and each test sample includes is less than predetermined threshold, it is determined that training
The risk index computation model test passes afterwards.
In the technical solution of above-described embodiment, the risk after training can be tested by the test sample collection of risk subject
The practicability of index computation model.Wherein risk index computation model test passes after determining training illustrate that practicability is good
It is good, on the contrary, risk index computation model test failure after training, then again refer to risk after adjustable training sample
Number computation model is trained.
With continued reference to shown in Fig. 2, in step S230, according to the risk index of each risk subject, institute will be in
It states the risk subject in presumptive area and is divided into multiple classes, each class therein corresponds to a risk index grade；Or according to
The corresponding sample size of each class is obtained from least one class corresponding according to the sequence of risk index from high to low
The risk subject of quantity.
In one embodiment of the application, multiple classes can be multiple risk classifications of risk subject, such as high risk master
Body, low-risk main body etc..Risk index grade can be different risk index section.Risk index can characterize each risk
The risk situation of main body, so the risk subject that risk index can be belonged to the same risk index grade be divided into it is same
Class.Risk analysis can be risk subject sampling supervision or risk subject situation statistics etc..Referred to by the risk for dividing different
Number grade, accurately can be divided into multiple classes for risk subject, risk subject is accurately segmented according to degree of risk.Into
And targetedly careful monitoring can be carried out based on the risk subject after subdivision, effectively improve monitoring efficiency.
Since the technical solution of the embodiment of the present application is will to be in presumptive area according to the risk index of each risk subject
Interior risk subject is divided into multiple classes.In order to guarantee that risk subject is divided into the flexibility and efficiency of multiple classes, step S230
Risk subject in the presumptive area is divided into multiple classes by the middle risk index according to each risk subject
Process, comprising:
According to risk index section locating for the risk index of each risk subject, same risk index area will be in
Between risk subject be divided into same class, to obtain the multiple class.
Risk index section is preset interval index.Such as it when the score that risk index divides for 0-100, can incite somebody to action
In risk subject in presumptive area, risk index is that 90-100 points of risk subject is divided into the first kind, risk index 60-
89 points of risk subject is divided into the second class, and risk index 0-60 points of risk subject is divided into third class.
Since the technical solution of the embodiment of the present application is to carry out wind to the risk subject based on the risk index grade
Danger analysis is described below and how to be based on risk index grade to risk subject progress risk analysis, as shown in figure 4, step
The process for being carried out risk analysis in S230 to the risk subject based on the risk index grade, is included the following steps:
Step S420 extracts phase from least one class in the multiple class according to the risk index grade respectively
The risk subject for answering quantity, using as the risk subject sampled out；
Step S420 carries out risk analysis based on the risk subject sampled out.
In one embodiment of the application, the corresponding class of each risk index grade may include at least one risk master
Body, while the risk subject risk situation in the different corresponding classes of risk index grade is different.Risk index higher grade
The risk of apoplectic stroke danger main body is higher.The risk subject for extracting respective numbers respectively from least one class in multiple classes, can
So that the risk subject sampled out can precisely represent the risk subject in presumptive area on the whole, and then guarantee risk master
The accuracy of body supervision, effectively promotes risk subject supervisory efficiency.Wherein, respective numbers can be preset according to demand, example
The extraction quantity for such as setting high risk index ranking is more than the extraction quantity of low-risk index ranking.It is appreciated that at least one
Extracting respectively in class can be each class from multiple classes and extracts respectively；It is also possible to be divided from the part class in multiple classes
It does not extract, for example, not extracting risk subject from the class of low-risk index ranking.
In one embodiment of the application, according to risk index grade, from least one class in multiple classes respectively
Extract the process of the risk subject of respective numbers, comprising:
Obtain the corresponding sampling probability of each risk index grade；
According to the corresponding sampling probability of each risk index grade, extracted respectively from least one class corresponding
The risk subject of quantity.
The corresponding sampling probability of each risk index grade, for example, risk index grade is 90-100 points: sampling is general
Rate is 80%；Risk index grade is 60-89 points: sampling probability 50%；Risk index grade 0-60 points: sampling probability is
5%.Sampling probability can be obtained based on the associated sampling probability of preset risk index grade, can also by user according to
Actual conditions are flexibly set, such as are flexibly set by government department according to law-enforcing ranks' situation, number of the enterprise, task complexity
It is fixed.As a result, according to the corresponding sampling probability of each risk index grade, can neatly be obtained in sampling corresponding
Quantity.
Technical solution based on the above embodiment be according to the corresponding sampling probability of each risk index grade,
Extract the risk subject of respective numbers respectively from least one class.It is also proposed in embodiments herein how according to sampling
Probability extracts the scheme of the risk subject of respective numbers, specific as follows:
In one embodiment of the application, according to the corresponding sampling probability of each risk index grade, from
The risk subject of respective numbers is extracted at least one class respectively, comprising:
According to the corresponding sampling probability of each risk index grade, the corresponding pumping of each class is calculated
Sample quantity；
According to the corresponding sample size of each class, respective numbers are randomly selected respectively from least one class
Risk subject or the risk subject for obtaining respective numbers from least one class according to the sequence of risk index from high to low.
In the present embodiment, according to the corresponding sampling probability of each risk index grade and corresponding risk subject
The corresponding sample size of each class can be calculated in total number.Pumping can be effectively ensured by way of randomly selecting
The randomness of sample.
The risk subject of respective numbers, such as certain are obtained from least one class according to the sequence of risk index from high to low
It include risk subject A, B, C, D in a class, corresponding risk index is followed successively by 99,96,95,90, such is calculated corresponding
Quantity is 3, then extracts risk subject A, B, C that 3 risk indexs are 99,96,95.It can guarantee to extract in each class most in this way
Multiple risk subjects of high risk guarantee the reliability that risk subject extracts.
The technical solution of previous embodiment makes it possible to be believed by risk index computation model according to the correlation of risk subject
Breath accurately obtains corresponding risk index, is then subdivided into the risk subject of all presumptive areas according to risk index multiple
Then class extracts the risk subject of respective numbers from least one class in multiple classes respectively, so that the risk master extracted
Body can precisely represent the risk subject in presumptive area, and then guarantee the accuracy of risk subject supervision, effectively promotion risk
Main body supervisory efficiency.
In one embodiment of the application, risk analysis is carried out to the risk subject based on the risk index grade
Process, can be with:
According to the risk index grade, the risk master of objective attribute target attribute is obtained from least one class in the multiple class
Body；
The corresponding class of risk subject based on the objective attribute target attribute generates the objective attribute target attribute risk subject of the corresponding class
List.
The garment marketing enterprises of objective attribute target attribute such as venture business's all properties, Internet enterprises etc..It is obtained from a class
The risk subject of objective attribute target attribute is taken, after being associated with by risk index with risk subject mark, generates target category in each class
The list of the risk subject of property, can clearly analyze the risk distribution situation of objective attribute target attribute enterprise.
In one embodiment of the application, risk analysis is carried out to the risk subject based on the risk index grade
Process, can be with:
According to the risk index grade, corresponding monitor control index is set for target class；
According to the target class monitor control index accordingly, the corresponding monitoring of apoplectic stroke danger main body described in timing acquisition refers to
Target data.
Monitor control index is exactly such as sales volume, tax amount, report information.Target class can be risk index grade highest
Class, be also possible to risk index grade be higher than predetermined threshold class.The monitor control index of risk index higher grade class setting
Also more.According to the monitor control index of target class, the data of the corresponding index of risk subject are periodically obtained, it can be with by these data
Continue the case where precisely analyzing risk subject.Wherein, periodically can be January/time or 1 day/time etc..
It, can be in one embodiment of the application
Risk subject list is generated according to the risk index of each risk subject, each risk subject list corresponds to one
Risk index grade.
Risk subject list is to be stored with each risk subject information (such as title, mailbox message) and corresponding risk refers to
Several lists.The corresponding storage of risk subject information is arrived respective risk by the list for preestablishing multiple and different risk index grades
The list of index ranking can generate risk subject list, convenient for the subdivision management of risk subject information.
In one embodiment of the application, risk index computation model includes XGBoost algorithm model.
The training method of XGBoost algorithm model includes:
The sample training collection of risk subject is obtained, each training sample that the sample training is concentrated includes risk subject
Relevant information and for the risk subject calibration risk index；
By the objective function of the feature vector input XGBoost algorithm of the training sample, the objective function meter is obtained
The first-loss value of calculation is added to the first regression tree function in the objective function；
According to the first-loss value, the objective function for being followed successively by the XGBoost algorithm adds the second regression tree function,
Until the first regression tree function forecasting risk index calculated and the prediction wind of all second regression tree function calculating
The difference for the risk index that the sum of dangerous index is demarcated with the risk subject is less than predetermined threshold, wherein the forecasting risk refers to
The risk index for the risk subject that the sum of number is predicted for XGBoost algorithm model.
In one embodiment of the application, XGBoost algorithm (Extreme Gradient Boosting) is GBDT
A kind of efficiently to realize, it provides a gradient and improves frame, it be designed to provide one " it is expansible, portable and
The gradient that can be distributed improves library ".XGBoost calculates scene with good using tree-model is promoted, to the risk index of risk subject
Good adaptability can provide better fitting result, form accurately risk index.
XGBoost algorithm model includes the regression tree after multiple optimizations, and each regression tree includes multiple leaf nodes, each
Leaf node corresponds to an index.After the feature vector of the risk information of risk subject is inputted XGBoost algorithm model, each
Regression tree according to the feature of input by traversal feature cut-off (such as be divided into left subtree when a feature vector is less than A, when
Right subtree is divided into when greater than A) risk subject is divided into a leaf node.Risk in each regression tree available in this way
The corresponding index of the corresponding leaf node of main body, the sum of index of all leaf nodes are exactly the prediction of XGBoost algorithm model
The risk index of risk subject.
Objective function is expression regression tree well-formed degree, the function of modelling effect in XGBoost algorithm.Objective function
Value it is smaller, return tree construction it is better, the prediction effect of model is better.
Objective function are as follows:
The objective function is made of two parts, first partFor measuring predictive index and true index
Gap, wherein i indicate i-th of sample, yiFor each training sample calibration risk index,For each training sample
Forecasting risk index,Indicate the prediction error (such as difference error) of i-th of sample.
Second partIt is then regularization term.Regularization term equally includes two parts, and T indicates leaf knot
The number of point, w indicate the index of leaf node.γ can control the number of leaf node, and λ can control the score of leaf node
It is not too big, over-fitting is prevented, all parameters can be added by mapping table.
Forecasting risk index calculates function in objective functionWherein, F represents all possible
Regression tree (regression tree function), fk(xi) represent regression tree function.As k=1, it is added to first regression tree (the first regression tree
Function)；When k=n, represents and be added to n regression tree in objective function altogether.
Wherein, fk(xi)=wq (x), w ∈ RT,q:Rd- > { 1,2 ..., T } indicates the set of all regression trees, and one is returned
Gui Shuyou T leaf node, it is a mapping that the value of this T leaf node, which constitutes T dimensional vector a w, q (x), is used to wind
Dangerous main body training sample is mapped to 1 and arrives some value of T, that is, risk subject is divided into some leaf node, and q (x) is in fact
Just represent the structure of CART tree.Wq (x) is exactly that wherein (risk subject is divided into certain to predicted value of the regression tree to sample x
The index of a leaf node).
Regression tree fk(xi) addition process, that is, objective function addition training process, by by the feature of training sample to
The objective function of the XGBoost algorithm of amount input the first regression tree function of addition obtains the first damage that the objective function calculates
Mistake value, while the first forecasting risk index of training sample that available first regression tree function calculates；Then successively add
Add the second regression tree function, each second regression tree is different regression tree.Each regression tree and all regression trees before
Training it is related to prediction.
When adding regression tree function according to first-loss value, need to guarantee in all regression tree functions added before
On the basis of, the every regression tree added later makes the value of objective function minimum.When all regression tree functions be calculated it is pre-
The sum for surveying risk index then determines regression tree (regression tree function) one shared k of addition less than predetermined threshold.Every addition one
The automatic re-optimization ginseng of result before the objective function meeting basis of regression tree function XGBoost algorithm on the basis of before
Number obtains the smallest XGBoost of objective function and promotes tree-model, and then realizes the XGBoost algorithm model of gradient optimizing, can be with
Guarantee the accuracy of risk index prediction.
Technical solution based on previous embodiment, in one embodiment of the application, the relevant information includes following
Combination any one or more: main body public feelings information, main body operation information, major network platform information and primary influences force information；
Wherein, the main body public feelings information is used to indicate the public feelings information of risk subject, and the main body operation information is used for
Indicate relevant information caused by the business activities of risk subject, the major network platform information is for indicating that risk subject is closed
The relevant information of the network platform of connection, the primary influences force information are used to indicate the related letter of the associated crowd of risk subject
Breath.
In one embodiment of the application, main body public feelings information can be included at least: the negative public sentiment details of network: negative
The specifying information of public sentiment；The negative public sentiment temperature of network: according to Internet communication range and duration；Customer complaint report: pass through
Complaints and denunciation small routine, public platform report center, etc. enterprise's complaints and denunciation information in data sources.Main body operation information can be down to
Less include: operation abnormal conditions: internet is disclosed to manage exception information；Abnormal selective examination record: it is taken out extremely disclosed in internet
Come to an end fruit and record；It pays taxes grade: register information of paying taxes disclosed in internet；Administrative penalty situation: administration disclosed in internet
Punish information；Executed person record: enterprise management level disclosed in internet is performed record information；Tax arrear record: internet is public
Tax arrear, the tax dodging, tax evasion information opened；It recruits situation: carrying out the information of personnel recruitment on internet.Major network platform information can
At least to include: network address IP situation: (black production refers to using internet as medium, with net for the corresponding IP information of main body network address and black production
Network technology is main means, and the illegal row of potential threat is brought for computer information system safety and cyberspace management order etc.
For) IP corresponding relationship；Customer service QQ, public platform, phone risk information: false propaganda, sham publicity；Main body mailbox risk: main body
Whether used mailbox is accused of false propaganda, sham publicity；Web page contents validity: main body business activities web page contents are true
Property, legality message；Whether web page contents publicize in violation of rules and regulations: the webpage of business operation whether violation information；Survival condition: public
Crowd number, various schools of thinkers number, webpage enliven situation information and flow information.Primary influences force information includes at least: the affiliated industry of main body
Influence power；Business impact range (such as association crowd's distributional region), number situation；Network attention situation；Customer service temperature changes feelings
Condition (such as network service temperature evaluation).
These information have magnanimity, the true, feature that is easily obtained from internet big data resource, can not depend on
Coordination and information sharing in many ways, obtains practical, high-efficient in practical work in information access process.Carrying out enterprise
Relevant information obtain when, can quickly, Overall Acquisition internet big data resource, guarantee cover large-scale, medium-sized, little Wei enterprise.
Below by taking enterprise's sampling system under a kind of application scenarios is to enterprise's sampling as an example, to the technical side of the embodiment of the present application
Case is described in detail.
Enterprise's sampling refers to: in routine monitoring selective examination work, by randomly selecting check object, selecting and appointing examination of law enforcement at random
Personnel spot-check situation and investigate and prosecute result in time to society's disclosure, guarantee the just, fair and open progress of supervision.In order to guarantee to supervise
The accuracy and promotion supervisory efficiency of pipe construct enterprise's sampling system, sample for enterprise, are realized and looked forward to based on the enterprise sampled out
Industry is precisely supervised.
In correlation technique, the check object of randomly selecting in enterprise's sampling process is emphasis therein, due in supervision work
Random chance, power are adjusted either using the true random in mathematical meaning or dependent on social public credit system in work
Strive realization precisely supervision, promotion supervisory efficiency.But the data of social public credit system acquisition are because coordinate that difficulty is big, timeliness
Property lower, coverage area focus primarily upon large and medium-sized enterprise, to current network economy, it is shared economical flourish, medium-sized and small enterprises
The case where increasing number bad adaptability.
Fig. 5 diagrammatically illustrates the process of the risk analysis method of the risk subject of one embodiment according to the application
Figure.
As shown in figure 5, acquiring data source, the i.e. enterprise from internet acquisition presumptive area first when carrying out enterprise's sampling
The relevant information of industry, including enterprise's public feelings information, enterprise operation information, enterprise network platform information, business impact force information.
Then, it is calculated by XGBoost algorithm model: being counted by the relevant information of the enterprise to above-mentioned presumptive area
It calculates, may include the training of XGBoost algorithm model in the part, the test verifying of XGBoost algorithm model obtains test verifying and closes
The XGBoost algorithm model of lattice, the step of prediction by the XGBoost algorithm model of test passes, final calculate generates enterprise
Risk index.
Finally, in enterprise's sampling system: the risk index list of enterprise can be generated, by business risk index list apoplexy
The high high probability sampling observation list (Enterprise Lists that such as business risk index is 80-100 points) as enterprise's sampling of dangerous index, it is low
Venture business's list is as low probability sampling observation list (Enterprise Lists that such as business risk index is 0-79 points).And then it can be never
The enterprise of respective numbers is extracted in same list.
Technical solution shown in fig. 5, by being believed in the public sentiment at internet end, operation, the network platform, influence power etc. enterprise
Breath is acquired, and can quickly, effectively cover the risk information of all kinds of enterprises, is carried out by XGBoost algorithm to above-mentioned data
Business risk index is calculated, the index can be used as according to predetermined regional enterprise is randomly selected the probability of supervised entities into
Row refinement, the low enterprise of probability, reduction risk index that extracts for improving the high enterprise of risk index extracts probability, to realize essence
Quasi- supervision promotes supervisory efficiency.
XGBoost algorithm model calculation risk index in application scenarios shown in fig. 5 is described in detail below.
Fig. 6 diagrammatically illustrates the process of the risk analysis method of the risk subject of one embodiment according to the application
Figure.
As shown in fig. 6, XGBoost algorithm model parameter setting may include: to be set by dictionary table or key-value pair list
All kinds of parameters (number, the number of features of tree of such as characteristics tree) during machine learning algorithm.
The training of XGBoost algorithm model may include: the sample training collection by obtaining enterprise, and the sample training is concentrated
Each training sample include enterprise relevant information and expert be directed to the enterprise calibration risk index；By the sample
Training sample in training set, which inputs in the risk index computation model, is trained the risk index computation model, leads to
Adjusting parameter is crossed, so that the risk index of each training sample of risk index computation model output and each training
The difference between risk index that sample includes is less than predetermined threshold.Then, the test sample collection of enterprise is obtained, the sample is surveyed
Each test sample that examination is concentrated includes the relevant information of enterprise and expert is directed to the risk index that the enterprise demarcates；By institute
The risk index is calculated in the risk index computation model after stating the test sample input training of test sample concentration
Model is tested, wherein if the risk index of each test sample of the risk index computation model output after training
The difference between risk index for including with each test sample is less than predetermined threshold, it is determined that the risk after training
It is qualified that index calculates model measurement.It wherein, can be with training for promotion effect by expert's notation methods.
For example, be for the sample [relevant information feature vector -> risk index] of an enterprise [(5,4,23,6) ->
100], XGBoost algorithm model is inputtedWherein parameter T indicates leaf knot
The number of point, w indicate that the index of leaf node, γ can control the number of leaf node, and λ can control the score of leaf node
It is not too big, over-fitting is prevented, all parameters can be added by mapping table.yiFor the risk index of sample calibration
It is 100,It is regression tree in risk forecast model to the forecasting risk index of sample.At this point, first regression tree of additionThe
It is 40 that one regression tree, which obtains forecasting risk index to this sample training, then obtain the forecasting risk index of first regression tree with
Demarcating risk index difference is 60.Then second regression tree is addedInput when so second regression tree training, the sample
Originally become [(5,4,23,6) -> 60], that is to say, that next regression tree tree input sample can be with the instruction of front decision tree
White silk is related to prediction, if the predicted value of second regression tree is 30, the forecasting risk index of the sample at this timeWherein, every posterior regression tree is in addition, it is ensured that in regression trees all before
On the basis of make XGBoost algorithm model functional value minimum l(t), guarantee modelling effect.Iteration adds regression tree, until working as
In XGBoost algorithm modelDifference is less than predetermined threshold (such as 99.95-100 > -0.1), the model training knot of the sample
Beam.
XGBoost algorithm model application may include: that qualified XGBoost algorithm model is exported after model training, use
In calculating business risk index.
The prediction of XGBoost algorithm model may include: according to XGBoost algorithm model, to making a reservation for local all enterprises
Relevant information carry out risk index calculating, the risk index of each enterprise is obtained, wherein each enterprise can have 0-100's
Risk index, index is higher, degree of risk is higher.
Technical solution shown in fig. 6, after input data (such as training sample data), by adjusting parameter, training XGBoost
Algorithm model obtains trained XGBoost algorithm model, by trained XGBoost algorithm model to input data (such as
Company-related information) risk index of enterprise is calculated, enterprise's accurately risk index can be obtained.
Fig. 7 shows the signal of the terminal interface of the risk analysis method of the risk subject applied to the embodiment of the present application
Figure.
It, can be with as shown in fig. 7, include selective examination object range setting regions in this exemplary embodiment, on terminal interface
Region (presumptive area as shown in Figure 2) belonging to selective examination enterprise is set as enterprise, the whole city；Type selection area is spot-check, choosing is passed through
Fixed " orientation " type has selected and carries out business risk analysis by the risk analysis method of risk subject shown in Fig. 2；Orientation
Scope region includes: high risk enterprise (enterprise that such as risk index is 90-100 points), (such as risk index is for medium risk enterprise
60-89 points of enterprise) and low-risk enterprise (enterprise that such as risk index is 0-59 points), by selecting high risk enterprise
To be to extract the enterprise of respective numbers in advanced enterprise from risk index grade；Selective examination ratio setting regions can be selected
High risk enterprise set sampling probability, when setting sampling probability as 80%, it can from being classified as taking out in high risk enterprise
The enterprise for taking wherein 80%, as the enterprise sampled out.It wherein, further include time started and end time setting regions, it can be with
The period of company-related information acquisition is set, realizes and obtains the related letter of the enterprise in presumptive area within a predetermined period of time
Breath.In interface shown in Fig. 7, by clicking the plan of preservation, it can extract 80% from the high risk enterprise in enterprise, the whole city
Enterprise.
Fig. 8 diagrammatically illustrates the block diagram of the risk analysis device of the risk subject of one embodiment according to the application.
As shown in figure 8, according to the risk analysis device 800 of the risk subject of one embodiment of the application, comprising: obtain
Module 810, prediction module 820 and analysis module 830.
Wherein, the relevant information that module 810 is used to obtain the risk subject in presumptive area is obtained；Prediction module
820 for inputting risk index computation model for the relevant information of the risk subject, and the risk for obtaining each risk subject refers to
Number；Analysis module 830 is used for the risk index according to each risk subject, by the risk master in the presumptive area
Body is divided into multiple classes, and each class therein corresponds to a risk index grade, to be based on the risk index grade to institute
It states risk subject and carries out risk analysis.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is used for,
According to the risk index grade, the wind of respective numbers is extracted respectively from least one class in the multiple class
Dangerous main body, using as the risk subject sampled out；
Risk analysis is carried out based on the risk subject sampled out.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is used for, and obtains each risk
The corresponding sampling probability of index ranking；According to the corresponding sampling probability of each risk index grade, from least
The risk subject of respective numbers is extracted in one class respectively.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is used for, according to each risk
The corresponding sampling probability of index ranking calculates the corresponding sample size of each class；According to each class point
Not corresponding sample size, randomly selects the risk subject of respective numbers respectively from least one class；Or according to described each
The corresponding sample size of class obtains the wind of respective numbers according to the sequence of risk index from high to low from least one class
Dangerous main body.
In some embodiments of the present application, aforementioned schemes are based on, the acquisition module is used for, and acquisition is in described predetermined
The risk analysis data of risk subject in region；By the data normalization for each attribute for including in the risk analysis data
For risk directional information, the risk directional information be used to indicate each attribute data whether Yi Chang information；According to
The risk directional information generates the relevant information of the risk subject.
In some embodiments of the present application, aforementioned schemes are based on, the acquisition module is used for, and is obtained in described predetermined
The relevant information of the risk subject in region within a predetermined period of time.
In some embodiments of the present application, aforementioned schemes are based on, the risk analysis device of the risk subject also wraps
Include: training module, for obtaining the sample training collection of risk subject, each training sample that the sample training is concentrated includes
The relevant information of risk subject and the risk index demarcated for the risk subject；The training sample that the sample training is concentrated
The risk index computation model is trained in this input risk index computation model, so that the risk index meter
The difference between risk index that the risk index and each training sample for calculating each training sample of model output include
Less than predetermined threshold.
In some embodiments of the present application, aforementioned schemes are based on, the risk analysis device of the risk subject also wraps
Include: test module, for obtaining the test sample collection of risk subject, each test sample that the test sample is concentrated includes
The relevant information of risk subject and the risk index demarcated for the risk subject；The test specimens that the test sample is concentrated
The risk index computation model is tested in the risk index computation model after this input training, wherein if instruction
The risk index of each test sample of risk index computation model output after white silk includes with each test sample
Risk index between difference be less than predetermined threshold, it is determined that training after the risk index computation model test passes.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is used for, according to each risk
Risk subject in same risk index section is divided into same by risk index section locating for the risk index of main body
Class, to obtain the multiple class.
In some embodiments of the present application, aforementioned schemes are based on, the analysis module is used for, according to each risk subject
Risk index generate risk subject list, each risk subject list correspond to a risk index grade.
In some embodiments of the present application, aforementioned schemes are based on, the risk index computation model includes that XGBoost is calculated
Method model.
The training module, for obtaining the sample training collection of risk subject, each training that the sample training is concentrated
Sample includes the relevant information of risk subject and the risk index for risk subject calibration；
By the objective function of the feature vector input XGBoost algorithm of the training sample, the objective function meter is obtained
The first-loss value of calculation is added to the first regression tree function in the objective function；
According to the first-loss value, the objective function for being followed successively by the XGBoost algorithm adds the second regression tree function,
Until the first regression tree function forecasting risk index calculated and the prediction wind of all second regression tree function calculating
The difference for the risk index that the sum of dangerous index is demarcated with the risk subject is less than predetermined threshold, wherein the forecasting risk refers to
The risk index for the risk subject that the sum of number is predicted for XGBoost algorithm model.
In some embodiments of the present application, aforementioned schemes are based on, the relevant information includes following any one or more
Combination: main body public feelings information, main body operation information, major network platform information and primary influences force information；Wherein, the main body
Public feelings information is used to indicate the public feelings information of risk subject, and the main body operation information is used to indicate the business activities of risk subject
Generated relevant information, the major network platform information are used to indicate the related letter of the associated network platform of risk subject
Breath, the primary influences force information are used to indicate the relevant information of the associated crowd of risk subject.
It should be noted that although being referred to several modules or list for acting the equipment executed in the above detailed description
Member, but this division is not enforceable.In fact, according to presently filed embodiment, it is above-described two or more
Module or the feature and function of unit can embody in a module or unit.Conversely, an above-described mould
The feature and function of block or unit can be to be embodied by multiple modules or unit with further division.
Fig. 9 shows the structural schematic diagram for being suitable for the computer system for the electronic equipment for being used to realize the embodiment of the present application.
It should be noted that the computer system 900 of the electronic equipment shown in Fig. 9 is only an example, it should not be to this Shen
Please embodiment function and use scope bring any restrictions.
As shown in figure 9, computer system 900 includes central processing unit (CPU) 901, it can be read-only according to being stored in
Program in memory (ROM) 902 or be loaded into the program in random access storage device (RAM) 903 from storage section 908 and
Execute various movements appropriate and processing.In RAM 903, it is also stored with various programs and data needed for system operatio.CPU
901, ROM 902 and RAM 903 is connected with each other by bus 904.Input/output (I/O) interface 905 is also connected to bus
904。
I/O interface 905 is connected to lower component: the importation 906 including keyboard, mouse etc.；It is penetrated including such as cathode
The output par, c 907 of spool (CRT), liquid crystal display (LCD) etc. and loudspeaker etc.；Storage section 908 including hard disk etc.；
And the communications portion 909 of the network interface card including LAN card, modem etc..Communications portion 909 via such as because
The network of spy's net executes communication process.Driver 910 is also connected to I/O interface 905 as needed.Detachable media 911, such as
Disk, CD, magneto-optic disk, semiconductor memory etc. are mounted on as needed on driver 910, in order to read from thereon
Computer program be mounted into storage section 908 as needed.
Particularly, according to an embodiment of the present application, it may be implemented as computer below with reference to the process of flow chart description
Software program.For example, embodiments herein includes a kind of computer program product comprising be carried on computer-readable medium
On computer program, which includes the program code for method shown in execution flow chart.In such reality
It applies in example, which can be downloaded and installed from network by communications portion 909, and/or from detachable media
911 are mounted.When the computer program is executed by central processing unit (CPU) 901, executes and limited in the system of the application
Various functions.
It should be noted that computer-readable medium shown in the application can be computer-readable signal media or meter
Calculation machine readable storage medium storing program for executing either the two any combination.Computer readable storage medium for example can be --- but not
Be limited to --- electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor system, device or device, or any above combination.Meter
The more specific example of calculation machine readable storage medium storing program for executing can include but is not limited to: have the electrical connection, just of one or more conducting wires
Taking formula computer disk, hard disk, random access storage device (RAM), read-only memory (ROM), erasable type may be programmed read-only storage
Device (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory device,
Or above-mentioned any appropriate combination.In this application, computer readable storage medium can be it is any include or storage journey
The tangible medium of sequence, the program can be commanded execution system, device or device use or in connection.And at this
In application, computer-readable signal media may include in a base band or as carrier wave a part propagate data-signal,
Wherein carry computer-readable program code.The data-signal of this propagation can take various forms, including but unlimited
In electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be that computer can
Any computer-readable medium other than storage medium is read, which can send, propagates or transmit and be used for
By the use of instruction execution system, device or device or program in connection.Include on computer-readable medium
Program code can transmit with any suitable medium, including but not limited to: wireless, electric wire, optical cable, RF etc. are above-mentioned
Any appropriate combination.
Flow chart and block diagram in attached drawing are illustrated according to the system of the various embodiments of the application, method and computer journey
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part of one module, program segment or code of table, a part of above-mentioned module, program segment or code include one or more
Executable instruction for implementing the specified logical function.It should also be noted that in some implementations as replacements, institute in box
The function of mark can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are practical
On can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it wants
It is noted that the combination of each box in block diagram or flow chart and the box in block diagram or flow chart, can use and execute rule
The dedicated hardware based systems of fixed functions or operations is realized, or can use the group of specialized hardware and computer instruction
It closes to realize.
Being described in unit involved in the embodiment of the present application can be realized by way of software, can also be by hard
The mode of part realizes that described unit also can be set in the processor.Wherein, the title of these units is in certain situation
Under do not constitute restriction to the unit itself.
As on the other hand, present invention also provides a kind of computer-readable medium, which be can be
Included in electronic equipment described in above-described embodiment；It is also possible to individualism, and without in the supplying electronic equipment.
Above-mentioned computer-readable medium carries one or more program, when the electronics is set by one for said one or multiple programs
When standby execution, so that the electronic equipment realizes method described in above-described embodiment.
It should be noted that although being referred to several modules or list for acting the equipment executed in the above detailed description
Member, but this division is not enforceable.In fact, according to presently filed embodiment, it is above-described two or more
Module or the feature and function of unit can embody in a module or unit.Conversely, an above-described mould
The feature and function of block or unit can be to be embodied by multiple modules or unit with further division.
Through the above description of the embodiments, those skilled in the art is it can be readily appreciated that example described herein is implemented
Mode can also be realized by software realization in such a way that software is in conjunction with necessary hardware.Therefore, according to the application
The technical solution of embodiment can be embodied in the form of software products, which can store non-volatile at one
Property storage medium (can be CD-ROM, USB flash disk, mobile hard disk etc.) in or network on, including some instructions are so that a calculating
Equipment (can be personal computer, server, touch control terminal or network equipment etc.) is executed according to the application embodiment
Method.
Those skilled in the art will readily occur to the application after considering specification and practicing embodiment disclosed herein
Other embodiments.This application is intended to cover any variations, uses, or adaptations of the application, these modifications are used
Way or adaptive change follow the application general principle and including the application it is undocumented in the art known in
Common sense or conventional techniques.
It should be understood that the application is not limited to the precise structure that has been described above and shown in the drawings, and
And various modifications and changes may be made without departing from the scope thereof.Scope of the present application is only limited by the accompanying claims.
Claims (15)
1. a kind of risk analysis method of risk subject characterized by comprising
Obtain the relevant information of the risk subject in presumptive area；
The relevant information of the risk subject is inputted into risk index computation model, obtains the risk index of each risk subject；
According to the risk index of each risk subject, the risk subject in the presumptive area is divided into multiple
Class, each class therein correspond to a risk index grade, with based on the risk index grade to the risk subject into
Row risk analysis.
2. the method according to claim 1, wherein described be based on the risk index grade to the risk master
Body carries out risk analysis, comprising:
According to the risk index grade, the risk master of respective numbers is extracted respectively from least one class in the multiple class
Body, using as the risk subject sampled out；
Risk analysis is carried out based on the risk subject sampled out.
3. according to the method described in claim 2, it is characterized in that, described according to the risk index grade, from the multiple
The risk subject of respective numbers is extracted at least one class in class respectively, comprising:
Obtain the corresponding sampling probability of each risk index grade；
According to the corresponding sampling probability of each risk index grade, respective numbers are extracted respectively from least one class
Risk subject.
4. according to the method described in claim 3, it is characterized in that, described respectively correspond according to each risk index grade
Sampling probability, extract the risk subject of respective numbers respectively from least one class, comprising:
According to the corresponding sampling probability of each risk index grade, the corresponding sampling number of each class is calculated
Amount；
According to the corresponding sample size of each class, the risk of respective numbers is randomly selected respectively from least one class
Main body；Or according to the corresponding sample size of each class, according to the sequence of risk index from high to low from least one
The risk subject of respective numbers is obtained in class.
5. the method according to claim 1, wherein the phase for obtaining the risk subject in presumptive area
Close information, comprising:
The risk analysis data of risk subject of the acquisition in the presumptive area；
It is risk directional information in the risk analysis data by the data normalization for each attribute for including, the risk is directed toward
Information be used for indicate each attribute data whether Yi Chang information；
The relevant information of the risk subject is generated according to the risk directional information.
6. the method according to claim 1, wherein the phase for obtaining the risk subject in presumptive area
Close information, comprising:
Obtain the relevant information of the risk subject in the presumptive area within a predetermined period of time.
7. the method according to claim 1, wherein the method also includes:
The sample training collection of risk subject is obtained, each training sample that the sample training is concentrated includes the phase of risk subject
Close information and the risk index for risk subject calibration；
The training sample that the sample training is concentrated is inputted in the risk index computation model and is calculated the risk index
Model is trained, so that the risk index of each training sample of risk index computation model output and each instruction
The difference practiced between the risk index that sample includes is less than predetermined threshold.
8. the method according to the description of claim 7 is characterized in that the method also includes:
The test sample collection of risk subject is obtained, each test sample that the test sample is concentrated includes the phase of risk subject
Close information and the risk index for risk subject calibration；
To the risk in the risk index computation model after the test sample input training that the test sample is concentrated
Index computation model is tested, wherein if each test sample of the risk index computation model output after training
The difference between risk index that risk index and each test sample include is less than predetermined threshold, it is determined that after training
The risk index computation model test passes.
9. the method according to claim 1, wherein the risk index according to each risk subject,
Risk subject in the presumptive area is divided into multiple classes, comprising:
According to risk index section locating for the risk index of each risk subject, same risk index section will be in
Risk subject is divided into same class, to obtain the multiple class.
10. the method according to claim 1, wherein being inputted in the relevant information by the risk subject
Risk index computation model, after obtaining the risk index of each risk subject, the method also includes:
Risk subject list is generated according to the risk index of each risk subject, each risk subject list corresponds to a risk
Index ranking.
11. the method according to claim 1, wherein the risk index computation model includes XGBoost algorithm
Model；
The method also includes:
The sample training collection of risk subject is obtained, each training sample that the sample training is concentrated includes the phase of risk subject
Close information and the risk index for risk subject calibration；
By the objective function of the feature vector input XGBoost algorithm of the training sample, obtain what the objective function calculated
First-loss value is added to the first regression tree function in the objective function；
According to the first-loss value, the objective function for being followed successively by the XGBoost algorithm adds the second regression tree function, until
The forecasting risk that the forecasting risk index and all second regression tree functions that the first regression tree function calculates calculate refers to
Number the sum of and the risk subject calibration risk index difference be less than predetermined threshold, wherein the forecasting risk index it
With the risk index for the risk subject predicted for XGBoost algorithm model.
12. method according to any one of claim 1 to 11, which is characterized in that the relevant information includes following
One or more of combination: main body public feelings information, main body operation information, major network platform information and primary influences force information；
Wherein, the main body public feelings information is used to indicate the public feelings information of risk subject, and the main body operation information is for indicating
Relevant information caused by the business activities of risk subject, the major network platform information is for indicating that risk subject is associated
The relevant information of the network platform, the primary influences force information are used to indicate the relevant information of the associated crowd of risk subject.
13. a kind of risk analysis device of risk subject characterized by comprising
Module is obtained, for obtaining the relevant information for the risk subject being in presumptive area；
Prediction module obtains each risk master for the relevant information of the risk subject to be inputted risk index computation model
The risk index of body；
Analysis module, for the risk index according to each risk subject, by the risk master in the presumptive area
Body is divided into multiple classes, and each class therein corresponds to a risk index grade, to be based on the risk index grade to institute
It states risk subject and carries out risk analysis.
14. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
Claim 1-12 described in any item methods are realized when being executed by processor.
15. a kind of electronic equipment characterized by comprising
Processor；And
Memory, for storing the computer program of the processor；Wherein, the processor is configured to via the execution meter
Calculation machine program carrys out perform claim and requires the described in any item methods of 1-12.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910678071.XA CN110458425A (en) | 2019-07-25 | 2019-07-25 | Risk analysis method, device, readable medium and the electronic equipment of risk subject |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910678071.XA CN110458425A (en) | 2019-07-25 | 2019-07-25 | Risk analysis method, device, readable medium and the electronic equipment of risk subject |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110458425A true CN110458425A (en) | 2019-11-15 |
Family
ID=68483459
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910678071.XA Pending CN110458425A (en) | 2019-07-25 | 2019-07-25 | Risk analysis method, device, readable medium and the electronic equipment of risk subject |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110458425A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110912749A (en) * | 2019-11-29 | 2020-03-24 | 北京工业大学 | Method for predicting DNS data |
| CN111614663A (en) * | 2020-05-20 | 2020-09-01 | 支付宝(杭州)信息技术有限公司 | Business risk determination method and device and electronic equipment |
| CN111859074A (en) * | 2020-07-29 | 2020-10-30 | 东北大学 | Internet public opinion information source influence assessment method and system based on deep learning |
| CN112101819A (en) * | 2020-10-28 | 2020-12-18 | 平安国际智慧城市科技股份有限公司 | Food risk prediction method, device, equipment and storage medium |
| CN112671744A (en) * | 2020-12-17 | 2021-04-16 | 杭州安恒信息技术股份有限公司 | Threat information processing method, device, equipment and storage medium |
| CN113971592A (en) * | 2021-12-23 | 2022-01-25 | 成都易播科技有限公司 | Supervision evaluation method, system and device for promotion information release main body |
| CN116629620A (en) * | 2023-07-26 | 2023-08-22 | 太平金融科技服务(上海)有限公司深圳分公司 | Risk level determining method and device, electronic equipment and storage medium |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103886217A (en) * | 2014-04-04 | 2014-06-25 | 江苏省环境科学研究院 | Ecological risk determining method for heavy metal pollution in river and lake sediments |
| CN104881783A (en) * | 2015-05-14 | 2015-09-02 | 中国科学院信息工程研究所 | E-bank account fraudulent conduct and risk detecting method and system |
| CN106611291A (en) * | 2016-12-13 | 2017-05-03 | 税云网络科技服务有限公司 | Information push method and device |
| CN107451710A (en) * | 2017-04-27 | 2017-12-08 | 北京鼎泰智源科技有限公司 | A kind of Information Risk grade five-category method and system |
| CN107993143A (en) * | 2017-11-23 | 2018-05-04 | 深圳大管加软件与技术服务有限公司 | A kind of Credit Risk Assessment method and system |
| CN109242280A (en) * | 2018-08-22 | 2019-01-18 | 泰康保险集团股份有限公司 | User behavior data processing method, device, electronic equipment and readable medium |
| CN109598931A (en) * | 2018-11-30 | 2019-04-09 | 江苏智通交通科技有限公司 | Group based on traffic safety risk divides and difference analysis method and system |
| CN109598414A (en) * | 2018-11-13 | 2019-04-09 | 阿里巴巴集团控股有限公司 | Risk evaluation model training, methods of risk assessment, device and electronic equipment |
| CN109829628A (en) * | 2019-01-07 | 2019-05-31 | 平安科技（深圳）有限公司 | Method for prewarning risk, device and computer equipment based on big data |
| CN109978680A (en) * | 2019-03-18 | 2019-07-05 | 杭州绿度信息技术有限公司 | A kind of air control method and system segmenting objective group's credit operation air control differentiation price |
-
2019
- 2019-07-25 CN CN201910678071.XA patent/CN110458425A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103886217A (en) * | 2014-04-04 | 2014-06-25 | 江苏省环境科学研究院 | Ecological risk determining method for heavy metal pollution in river and lake sediments |
| CN104881783A (en) * | 2015-05-14 | 2015-09-02 | 中国科学院信息工程研究所 | E-bank account fraudulent conduct and risk detecting method and system |
| CN106611291A (en) * | 2016-12-13 | 2017-05-03 | 税云网络科技服务有限公司 | Information push method and device |
| CN107451710A (en) * | 2017-04-27 | 2017-12-08 | 北京鼎泰智源科技有限公司 | A kind of Information Risk grade five-category method and system |
| CN107993143A (en) * | 2017-11-23 | 2018-05-04 | 深圳大管加软件与技术服务有限公司 | A kind of Credit Risk Assessment method and system |
| CN109242280A (en) * | 2018-08-22 | 2019-01-18 | 泰康保险集团股份有限公司 | User behavior data processing method, device, electronic equipment and readable medium |
| CN109598414A (en) * | 2018-11-13 | 2019-04-09 | 阿里巴巴集团控股有限公司 | Risk evaluation model training, methods of risk assessment, device and electronic equipment |
| CN109598931A (en) * | 2018-11-30 | 2019-04-09 | 江苏智通交通科技有限公司 | Group based on traffic safety risk divides and difference analysis method and system |
| CN109829628A (en) * | 2019-01-07 | 2019-05-31 | 平安科技（深圳）有限公司 | Method for prewarning risk, device and computer equipment based on big data |
| CN109978680A (en) * | 2019-03-18 | 2019-07-05 | 杭州绿度信息技术有限公司 | A kind of air control method and system segmenting objective group's credit operation air control differentiation price |
Non-Patent Citations (1)
| Title |
|---|
| 蔡文杰: "基于风险的电梯监督抽查管理体系及应用研究", 中国硕士学位论文全文数据库工程科技Ⅱ辑, no. 12, pages 24 - 90 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110912749A (en) * | 2019-11-29 | 2020-03-24 | 北京工业大学 | Method for predicting DNS data |
| CN111614663A (en) * | 2020-05-20 | 2020-09-01 | 支付宝(杭州)信息技术有限公司 | Business risk determination method and device and electronic equipment |
| CN111614663B (en) * | 2020-05-20 | 2022-04-08 | 杭州蚂蚁聚慧网络技术有限公司 | Business risk determination method and device and electronic equipment |
| CN111859074A (en) * | 2020-07-29 | 2020-10-30 | 东北大学 | Internet public opinion information source influence assessment method and system based on deep learning |
| CN111859074B (en) * | 2020-07-29 | 2023-12-29 | 东北大学 | Network public opinion information source influence evaluation method and system based on deep learning |
| CN112101819A (en) * | 2020-10-28 | 2020-12-18 | 平安国际智慧城市科技股份有限公司 | Food risk prediction method, device, equipment and storage medium |
| CN112101819B (en) * | 2020-10-28 | 2024-08-23 | 平安国际智慧城市科技股份有限公司 | Food risk prediction method, device, equipment and storage medium |
| CN112671744A (en) * | 2020-12-17 | 2021-04-16 | 杭州安恒信息技术股份有限公司 | Threat information processing method, device, equipment and storage medium |
| CN113971592A (en) * | 2021-12-23 | 2022-01-25 | 成都易播科技有限公司 | Supervision evaluation method, system and device for promotion information release main body |
| CN116629620A (en) * | 2023-07-26 | 2023-08-22 | 太平金融科技服务(上海)有限公司深圳分公司 | Risk level determining method and device, electronic equipment and storage medium |
| CN116629620B (en) * | 2023-07-26 | 2024-02-02 | 太平金融科技服务(上海)有限公司深圳分公司 | Risk level determining method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110458425A (en) | Risk analysis method, device, readable medium and the electronic equipment of risk subject | |
| CN110390465A (en) | Air control analysis and processing method, device and the computer equipment of business datum | |
| CN107908606A (en) | Method and system based on different aforementioned sources automatic report generation | |
| CN107689008A (en) | A kind of user insures the method and device of behavior prediction | |
| CN110008259A (en) | The method and terminal device of visualized data analysis | |
| US8793210B2 (en) | General market prediction using position specification language | |
| CN110084627A (en) | The method and apparatus for predicting target variable | |
| CN111199474A (en) | Risk prediction method and device based on network diagram data of two parties and electronic equipment | |
| CN111222976A (en) | Risk prediction method and device based on network diagram data of two parties and electronic equipment | |
| CN113254542A (en) | Data visualization processing method and device and electronic equipment | |
| CN109614319B (en) | Automatic testing method and device, electronic equipment and computer readable medium | |
| CN113298121A (en) | Message sending method and device based on multi-data source modeling and electronic equipment | |
| CN112990311A (en) | Method and device for identifying admitted client | |
| CN113762973A (en) | Data processing method and device, computer readable medium and electronic equipment | |
| CN112488865A (en) | Financial risk prediction method and device based on financial time nodes and electronic equipment | |
| CN116228402A (en) | Financial credit investigation feature warehouse technical support system | |
| CN116402546A (en) | Store risk attribution method and device, equipment, medium and product thereof | |
| CN109684844A (en) | A kind of webshell detection method and device | |
| US11972356B2 (en) | System and/or method for an autonomous linked managed semantic model based knowledge graph generation framework | |
| CN115238292A (en) | Data security management and control method and device, electronic equipment and storage medium | |
| CN111429257B (en) | Transaction monitoring method and device | |
| CN113902545A (en) | Resource limit distribution method and device and electronic equipment | |
| Pang et al. | Wt model & applications in loan platform customer default prediction based on decision tree algorithms | |
| Tan et al. | Temporal association rule mining | |
| CN112819472A (en) | Fraud prediction method and device for resource raising project and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |