CN106022915A - Enterprise credit risk assessment method and apparatus - Google Patents
Enterprise credit risk assessment method and apparatus Download PDFInfo
- Publication number
- CN106022915A CN106022915A CN201610515496.5A CN201610515496A CN106022915A CN 106022915 A CN106022915 A CN 106022915A CN 201610515496 A CN201610515496 A CN 201610515496A CN 106022915 A CN106022915 A CN 106022915A
- Authority
- CN
- China
- Prior art keywords
- factor
- objective
- conclusion
- analytical
- risk
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides an enterprise credit risk assessment method and an apparatus. The method comprises the steps of extracting an objective factor out of an objective circumstance survey report of a to-be-trusted enterprise, and adopting the objective factor of the objective circumstance survey report as a first objective factor; inputting the first objective factor into each pre-stored risk analysis model, and outputting a first analysis factor corresponding to each risk analysis model; inputting the first objective factor and the first analysis factor into each pre-stored risk judgment model and outputting a first conclusion factor corresponding to each risk judgment model; executing the operation till the outputted first conclusion factor does not change at all. The above operation comprises the steps of inputting the first objective factor, the first analysis factor and the first conclusion factor of the last output result into each pre-stored risk judgment model, outputting the first conclusion factor corresponding to each risk judgment model, and combining the first objective factor, the first analysis factor and all first conclusion factors to form a credit risk view.
Description
Technical field
The present embodiments relate to technical field of data processing, particularly relate to a kind of Enterprise Credit Risk Evaluation
Method and apparatus.
Background technology
Along with deepening continuously of market economy, bank credit scale constantly expands, and bank is to business loan
Time, need enterprise is carried out assessing credit risks, the enterprise that faithlessness risk or credit risk are relatively low is entered
Row behavior of lending.
What bank's loan prosomite was used at present carries out the method for assessing credit risks mainly with expert to enterprise
Empirical model is main, is aided with real data and is adjusted.So the assessment that relates to of expertise model because of
The subjectivity of element and index is relatively big, and is difficult to ensure that the comprehensive of assessment.Meanwhile, expertise model
Update cycle longer, update every time and be required to put into substantial amounts of manpower and materials, cause expertise model
Hysteresis quality the most obvious.
Summary of the invention
The embodiment of the present invention provides a kind of Enterprise Credit Risk Evaluation method, solves in prior art and passes through
Enterprise Credit Risk Evaluation is caused assessment result to have subjectivity, one-sidedness and delayed by expertise model
The problem of property.
The embodiment of the present invention provides a kind of Enterprise Credit Risk Evaluation method, including:
Extract the objective factor in the objective circumstances investigation report intending trusted enterprise, described plan trusted enterprise
The objective factor extracted in objective circumstances investigation report is the first objective factor；
Described first objective factor is input in each risk analysis model of pre-stored, export described often
The first analytical factor that individual risk analysis model is corresponding；
Described first objective factor, described first analytical factor are input to each risk judgment of pre-stored
In model, export the first conclusion factor that described each risk judgment model is corresponding；
Performing operation until the first conclusion factor of described operation output no longer changes, described operation is:
By described first objective factor, described first analytical factor, the first conclusion factor of described last output
It is input in each risk judgment model of described pre-stored, exports described each risk judgment model corresponding
The first conclusion factor；
Described first objective factor, described first analytical factor, the first whole conclusion factors are closed
And, form the credit risk view of described plan trusted enterprise；
Wherein, each risk analysis model of described pre-stored and each risk judgment mould of described pre-stored
Type is to build according to Duo Pian bank responsible investigation report at no distant date.
The embodiment of the present invention provides a kind of Enterprise Credit Risk Evaluation device, including:
Extraction module, for extracting the objective factor in the objective circumstances investigation report intending trusted enterprise, institute
The objective factor extracted in the objective circumstances investigation report of Shu Ni trusted enterprise is the first objective factor；
Output module, for being input to each risk analysis model of pre-stored by described first objective factor
In, export the first analytical factor that described each risk analysis model is corresponding；
Output module, is additionally operable to be input to prestore by described first objective factor, described first analytical factor
In each risk judgment model of storage, export the first conclusion factor that described each risk judgment model is corresponding；
Operation module, for performing operation until the first conclusion factor of described operation output is no longer changed to
Only, described operation is: by described first objective factor, described first analytical factor, described last time defeated
The the first conclusion factor gone out is input in each risk judgment model of described pre-stored, exports described each
The first conclusion factor that risk judgment model is corresponding；
Merge module, for by described first objective factor, described first analytical factor, whole first
Conclusion factor merges, and forms the credit risk view of described plan trusted enterprise；
Wherein, each risk analysis model of described pre-stored and each risk judgment mould of described pre-stored
Type is to build according to Duo Pian bank responsible investigation report at no distant date.
The embodiment of the present invention provides a kind of Enterprise Credit Risk Evaluation method and apparatus, intends trusted by extracting
Objective factor in the objective circumstances investigation report of enterprise, intends in the objective circumstances investigation report of trusted enterprise
The objective factor extracted is the first objective factor；First objective factor is input to each risk of pre-stored
Analyze in model, export the first analytical factor that each risk analysis model is corresponding；By the first objective factor,
First analytical factor is input in each risk judgment model of pre-stored, exports each risk judgment model
The first corresponding conclusion factor；Perform operation until the conclusion factor of operation output no longer changes, behaviour
As: the first objective factor, the first analytical factor, the first conclusion factor of last output are input to
In each risk judgment model of pre-stored, export the first conclusion factor that each risk judgment model is corresponding；
First objective factor, the first analytical factor, the first whole conclusion factors are merged, forms plan and be subject to
The credit risk view of letter enterprise；Each risk analysis model and each wind of pre-stored due to pre-stored
Danger judgment models is to build according to Duo Pian bank responsible investigation report at no distant date, so in Duo Pian bank
Responsible investigation report can include various situation, it is ensured that the objectivity of assessment and comprehensive.And can
According to the responsible investigation report of recent Duo Pian bank, model is updated at any time, it is to avoid model delayed
Property.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to reality
Execute the required accompanying drawing used in example or description of the prior art to be briefly described, it should be apparent that under,
Accompanying drawing during face describes is some embodiments of the present invention, for those of ordinary skill in the art,
On the premise of not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the flow chart of Enterprise Credit Risk Evaluation embodiment of the method one of the present invention；
Fig. 2 is the flow chart of Enterprise Credit Risk Evaluation embodiment of the method two of the present invention；
Fig. 3 is the example schematic using assessing credit risks method in the embodiment of the present invention two；
Fig. 4 is the structural representation of Enterprise Credit Risk Evaluation device embodiment one of the present invention；
Fig. 5 is the structural representation of Enterprise Credit Risk Evaluation device embodiment two of the present invention.
Detailed description of the invention
For making the purpose of the embodiment of the present invention, technical scheme and advantage clearer, below in conjunction with this
Accompanying drawing in bright embodiment, is clearly and completely described the technical scheme in the embodiment of the present invention,
Obviously, described embodiment is a part of embodiment of the present invention rather than whole embodiments.Based on
Embodiment in the present invention, those of ordinary skill in the art are obtained under not making creative work premise
The every other embodiment obtained, broadly falls into the scope of protection of the invention.
Should be appreciated that term "and/or" used herein is only a kind of association describing affiliated partner
Relation, can there are three kinds of relations, such as, A and/or B, can represent in expression: individualism A,
There is A and B, individualism B these three situation simultaneously.It addition, character "/", general table herein
Show the forward-backward correlation relation to liking a kind of "or".
Depend on linguistic context, word as used in this " if " can be construed to " ... time "
Or " when ... " or " in response to determining " or " in response to detection ".Similarly, depend on linguistic context,
Phrase " if it is determined that " or " if detection (condition of statement or event) " can be construed to " when
When determining " or " in response to determining " or " when detecting (condition of statement or event) " or " response
In detection (condition of statement or event) ".
Fig. 1 is the flow chart of Enterprise Credit Risk Evaluation embodiment of the method one of the present invention, as it is shown in figure 1,
The executive agent of the present embodiment is Enterprise Credit Risk Evaluation device, this Enterprise Credit Risk Evaluation device
May be located at the application of local terminal, or can also be located locally the plug-in unit in the application of terminal or
The functional units such as SDK (Software Development Kit, SDK), the present invention is real
Execute example this is not particularly limited.
It is understood that the application program (nativeApp) that application can be mounted in terminal, or
Person can also is that a web page program (webApp) of the browser in terminal, and the embodiment of the present invention is to this
It is not defined.
The Enterprise Credit Risk Evaluation method that then the present embodiment provides includes following step.
Step 101, extracts the objective factor in the objective circumstances investigation report intending trusted enterprise.
Wherein, intend trusted enterprise objective circumstances investigation report in extract objective factor be first objective because of
Element, the first objective factor can be multiple.
Specifically, in the present embodiment, intending trusted enterprise is to borrow prosomite to carry out the enterprise of assessing credit risks.
Include that the objective circumstances of trusted enterprise are carried out by all departments at the objective circumstances investigation report intending trusted enterprise
The result of investigation.First the objective circumstances investigation report intending trusted enterprise can be carried out word segmentation processing, then will divide
Word process after each participle and pre-stored the data base including whole objective factor in participle carry out
Joining, the participle matched extracts, it is thus achieved that multiple first objective factors in objective circumstances investigation report.
In the present embodiment, it is possible to after carrying out word segmentation processing, according to the structure of " noun phrase+numeral phrase ",
Extract the multiple participles meeting " noun phrase+numeral phrase " structure in objective circumstances investigation report,
Constitute corresponding multiple first objective factors.The first objective factor as extracted can be " accounts receivable
30000000 yuan ".
In the present embodiment, it is possible to use other mode to extract the objective circumstances investigation report intending trusted enterprise
In multiple first objective factors, the present embodiment does not limits.
Step 102, is input to the first objective factor in each risk analysis model of pre-stored, output
The first analytical factor that each risk analysis model is corresponding.
Wherein, the first analytical factor is each risk analysis mould that the first objective factor is input to pre-stored
In type, export the analytical factor that each risk analysis model is corresponding.
In the present embodiment, construct multiple risk previously according to Duo Pian bank responsible investigation report at no distant date
Analyzing model, the responsible investigation of Mei Pian bank is reported as borrowing the tune to corresponding trusted enterprise that prosomite is formed
Look into situation and Risk Assessment Report.Wherein, Duo Pian bank responsible investigation report at no distant date can be nearest
In 1 year or in nearest one month or the Duo Pian bank responsible investigation report of other recent times, the present embodiment
In this is not limited.
In the present embodiment, the first analytical factor can possess " noun phrase+adjective " structure.
Specifically, in the present embodiment, due to each risk analysis model characterize objective factor with analyze because of
The incidence relation of element, so being input to each risk analysis mould of pre-stored by the first whole objective factors
In type, by calculating, export the first analytical factor that each risk analysis model is corresponding.Wherein, output
The first analytical factor corresponding to each risk analysis model can be multiple.
Step 103, each risk that the first objective factor, the first analytical factor are input to pre-stored is sentenced
In disconnected model, export the first conclusion factor that each risk judgment model is corresponding.
Wherein, the first conclusion factor is that the first objective factor, the first analytical factor are input to pre-stored
In each risk judgment model, export the conclusion factor that each risk judgment model is corresponding.
Specifically, in the present embodiment, each risk judgment model of pre-stored is according to many at no distant date
Bank's responsible investigation report builds.Due to risk judgment model characterize objective factor, analytical factor with
The incidence relation of conclusion factor, so being input to pre-stored by the first objective factor, the first analytical factor
After in each risk judgment model, by calculating, export and the first objective factor, the first analytical factor phase
The first conclusion factor corresponding to each risk judgment model of association.Output with the first objective factor, the
The first conclusion factor that each risk judgment model that one analytical factor is associated is corresponding can be multiple.
Wherein, the first conclusion factor can possess " noun phrase+verb phrase " structure.As conclusion factor is
" enterprise's downstream returned money is slow ", " enterprise order is abnormal " etc..
Step 104, performs operation until the conclusion factor of operation output no longer changes, operates and be:
First objective factor, the first analytical factor, the first conclusion factor of last output are input to pre-stored
Each risk judgment model in, export the first conclusion factor that each risk judgment model is corresponding.
Specifically, in the present embodiment, the first objective factor, the first analytical factor, step 103 are exported
The first conclusion factor be input in each risk judgment model of pre-stored, export each risk judgment mould
The first conclusion factor that type is corresponding, in the first conclusion factor this exported and step 103 the of output
One conclusion factor contrasts, if the first conclusion factor of output changes, then continues executing with first
Objective factor, the first analytical factor, the first conclusion factor of last output are input to each of pre-stored
In risk judgment model, export the operation of the first conclusion factor corresponding to each risk judgment model, until
Till first conclusion factor of output no longer changes.
In the present embodiment, performed the first objective factor, the first analytical factor, last defeated by circulation
The the first conclusion factor gone out is input in each risk judgment model of pre-stored, exports each risk judgment
The operation of the first conclusion factor that model is corresponding, it is possible to obtain the first objective factor, the first analytical factor and
Each incidence relation between first conclusion factor, until the first conclusion factor that output is final.
Step 105, is carried out the first objective factor, the first analytical factor, the first whole conclusion factors
Merge, form the credit risk view intending trusted enterprise.
Specifically, in the present embodiment, formed and the credit risk view intending trusted enterprise may include that the
One objective factor, the first analytical factor, the first whole conclusion factors, and the first objective factor and the
The relation of one analytical factor, the first objective factor, the first analytical factor and the relation of the first conclusion factor,
Relation etc. between first conclusion factor.
In the present embodiment, by extracting the objective factor in the objective circumstances investigation report intending trusted enterprise,
The objective factor intending extracting in the objective circumstances investigation report of trusted enterprise is the first objective factor；By first
Objective factor is input in each risk analysis model of pre-stored, exports each risk analysis model corresponding
The first analytical factor；First objective factor, the first analytical factor are input to each risk of pre-stored
In judgment models, export the first conclusion factor that each risk judgment model is corresponding；Perform operation until grasping
Make till the conclusion factor that exports no longer changes, operate and be: by the first objective factor, the first analytical factor,
First conclusion factor of last output is input in each risk judgment model of pre-stored, exports each
The first conclusion factor that risk judgment model is corresponding；By the first objective factor, the first analytical factor, all
The first conclusion factor merge, formed intend trusted enterprise credit risk view；Due to pre-stored
Each risk judgment model of each risk analysis model and pre-stored is according to Duo Pian bank at no distant date to the greatest extent
Duty investigation report builds, so can include various situation in the responsible investigation report of Duo Pian bank, protects
Demonstrate,prove the objectivity and comprehensive of assessment.And can be at any time according to recent Duo Pian bank responsible investigation report
Model is updated, it is to avoid the hysteresis quality of model.
Fig. 2 is the flow chart of Enterprise Credit Risk Evaluation embodiment of the method two of the present invention, as in figure 2 it is shown,
The Enterprise Credit Risk Evaluation method that the present embodiment provides, compared to embodiment one, is one and is more highly preferred to
Embodiment, then the present embodiment provide Enterprise Credit Risk Evaluation method comprise the following steps.
Step 201, it may be judged whether storage has the risk analysis model built in advance and risk judgment model,
The most then perform step 207, if it is not, then perform step 202.
Specifically, in the present embodiment, during owing to plan trusted enterprise is carried out assessing credit risks, need
Use the risk analysis model and risk judgment model building in advance and storing, so to intending trusted enterprise
Before industry carries out assessing credit risks, it is judged that whether store the risk built in advance at default memory area
Analyze model and risk judgment model, the most then perform plan trusted enterprise is carried out assessing credit risks
Step, otherwise, then build risk analysis model and risk judgment model.
Step 202, extract the second objective factor in Duo Pian bank responsible investigation report at no distant date,
Second analytical factor and the second conclusion factor.
Wherein, the objective factor extracted from Duo Pian bank responsible investigation report at no distant date is the second visitor
Sight factor, the analytical factor extracted from Duo Pian bank responsible investigation report at no distant date is the second analysis
Factor, the conclusion factor extracted from Duo Pian bank responsible investigation report at no distant date be the second conclusion because of
Element.
Further, in step 202, extract the in Duo Pian bank responsible investigation report at no distant date
Two objective factors, the second analytical factor and the second conclusion factor specifically include:
According to the structure of " noun phrase+numeral phrase ", extract Duo Pian bank responsible investigation at no distant date
The second objective factor in report；According to the structure of " noun phrase+adjective ", extract at no distant date
Duo Pian bank responsible investigation report in the second analytical factor；According to " noun phrase+verb phrase "
Structure, extract the second conclusion factor in Duo Pian bank responsible investigation report at no distant date.
Wherein, numeral phrase refers to the noun phrase of word centered by numeral.
Specifically, in the present embodiment, owing to objective factor possesses the knot of " noun phrase+numeral phrase "
Structure, so, according to the structure of " noun phrase+numeral phrase ", extract Duo Pian bank at no distant date to the greatest extent
The second objective factor in duty investigation report.Analytical factor possesses the structure of " noun phrase+adjective ",
So according to the structure of " noun phrase+adjective ", extracting Duo Pian bank responsible investigation at no distant date
The second analytical factor in report.Conclusion factor possesses the structure of " noun phrase+verb phrase ".
So according to the structure of " noun phrase+verb phrase ", extracting the responsible tune of Duo Pian bank at no distant date
Look into the second conclusion factor in report.
In the present embodiment, it is also possible to first by all of objective factor known in advance, analytical factor,
Conclusion factor stores.Duo Pian bank responsible investigation report at no distant date will carry out word segmentation processing,
Participle after word segmentation processing is carried out with the objective factor of storage, analytical factor, conclusion factor respectively
Joining, extract the participle of coupling, the participle of this coupling is from Duo Pian bank responsible investigation at no distant date
The second objective factor extracted in report or the second analytical factor or the second conclusion factor.
Step 203, is carried out respectively to the second objective factor, the second analytical factor and the second conclusion factor
Cluster, forms objective factor collection, analytical factor collection, conclusion set of factors.
Specifically, in the present embodiment, can according to whether have identical information respectively to the second objective factor,
Second analytical factor and the second conclusion factor cluster, formed objective factor collection, analytical factor collection,
Conclusion set of factors.Concentrate in different objective factors, can have the second identical objective factor, i.e. visitor
Can have inclusion relation between sight set of factors and maybe can have common factor.In like manner, different analyses because of
Element is concentrated, and can have inclusion relation and maybe can have common factor.In different conclusion set of factors, can
Maybe can have common factor having inclusion relation.
Step 204, builds risk analysis decision-tree model according to objective factor collection and analytical factor collection.
Further, in the present embodiment, with whole objective factor collection for input, whole analytical factors
Collection is output, uses pruning algorithms or rear pruning algorithms in advance to build multiple risk analysis decision-tree models.
Specifically, in the present embodiment, with whole objective factor collection for input, whole analytical factor collection
For output, according to recent multiple in the responsible investigation report of Duo Pian bank in the second objective factor, the
The relation of two analytical factors, uses pruning algorithms or rear pruning algorithms in advance to build multiple risk analysis decision-makings
Tree-model.
Step 205, builds risk judgment certainly according to objective factor, analytical factor collection and conclusion set of factors
Plan tree-model.
Further, in the present embodiment, with whole objective factor collection, whole analytical factors with except making
Other conclusion set of factors outside for the conclusion set of factors of output are input, and multiple conclusion set of factors are output,
Use pruning algorithms or rear pruning algorithms in advance, build risk judgment decision-tree model.
Specifically, in the present embodiment, using whole objective factor collection, whole analytical factors and except as
Other conclusion set of factors outside the conclusion set of factors of output are input, and multiple conclusion set of factors are output, root
According to the second objective factor in Duo Pian bank responsible investigation report at no distant date, the second analytical factor, second
The relation of conclusion factor, uses pruning algorithms or rear pruning algorithms in advance to build multiple risk judgment decision trees
Model.Final election of laying equal stress on takes different multiple conclusion set of factors combination conduct output, according to said method, structure
Build multiple risk judgment decision-tree model, until not having new risk judgment decision-tree model to generate.
Step 206, by each risk analysis decision-tree model built and risk judgment decision-tree model
Store.
Specifically, in the present embodiment, each risk analysis decision-tree model built and risk can be sentenced
Disconnected decision-tree model stores respectively, is specifically storable in a certain storage chip or fixed storage region
In.
After execution of step 206, perform step 207.
Step 207, extracts the objective factor in the objective circumstances investigation report intending trusted enterprise.
Wherein, the objective factor intending extracting in the objective circumstances investigation report of trusted enterprise is first objective
Factor.
Step 208, is input to the first objective factor in each risk analysis model of pre-stored, defeated
Go out the first analytical factor that each risk analysis model is corresponding.
Step 209, is input to each risk of pre-stored by the first objective factor, the first analytical factor
In judgment models, export the first conclusion factor that each risk judgment model is corresponding.
Step 210, performs operation until the first conclusion factor of operation output no longer changes, behaviour
As: by the first objective factor, the first analytical factor, the first conclusion factor input of last output
In each risk judgment model of pre-stored, export the first conclusion that each risk judgment model is corresponding
Factor.
Step 211, enters the first objective factor, the first analytical factor, the first whole conclusion factors
Row merges, and forms the credit risk view intending trusted enterprise.
Specifically, in the present embodiment, the implementation of step 207-step 211 and the embodiment of the present invention one
In the implementation of step 101-step 105 identical, this is no longer going to repeat them.
In the present embodiment, by judging whether to store the risk analysis model built in advance and risk is sentenced
Disconnected model, if it is not, the second objective factor then extracted in Duo Pian bank responsible investigation report at no distant date,
Second analytical factor and the second conclusion factor, to the second objective factor, the second analytical factor and the second knot
Opinion factor clusters respectively, forms objective factor collection, analytical factor collection, conclusion set of factors, according to
Objective factor collection and analytical factor collection build risk analysis decision-tree model, according to objective factor, analysis
Set of factors and conclusion set of factors build risk judgment decision-tree model, each risk analysis built are determined
Plan tree-model and risk judgment decision-tree model store, then to the credit risk intending trusted enterprise
It is estimated, not only ensure that the objectivity of assessment, comprehensive, it is to avoid hysteresis quality, and improve
Bank borrow before the utilization rate of responsible investigation report, before borrowing due to bank, responsible investigation report has the highest
Accuracy, so according to Duo Pian bank responsible investigation report at no distant date build risk analysis model and
Risk judgment model, is estimated the credit risk intending trusted enterprise, also improves plan trusted enterprise
Carry out the accuracy rate of assessing credit risks.
Further, the risky decision making tree-model in the present embodiment formed said method illustrates.
Fig. 3 is the example schematic using assessing credit risks method in the embodiment of the present invention two.As it is shown on figure 3,
The first objective factor in Fig. 3 have respectively " accounts receivable remaining sum ... unit ", " balance of deposits ... unit ",
First conclusion is because have " accounts receivable amplification is bigger than normal/general ", " enterprise order is abnormal/normal ".By certain
After one risk judgment model, if " accounts receivable remaining sum is more than or equal to 20,000,000 yuan " and " accounts receivable increasing
Bigger than normal " the first corresponding conclusion factor of then exporting is " enterprise's downstream returned money is slow "；By a certain risk
After judgment models, if " accounts receivable remaining sum is more than or equal to 20,000,000 yuan " and " accounts receivable amplification is general "
And " enterprise order is abnormal ", then the first corresponding conclusion factor exported is " enterprise downstream exists abnormal "；
" if accounts receivable remaining sum is more than or equal to 20,000,000 yuan " and " accounts receivable is not that growth is fast " and " enterprise
Industry order is normal ", then the first corresponding conclusion factor exported is " enterprise is in the growth stage "；If " should
Receipt on account money remaining sum is less than 20,000,000 yuan " and " inventory balance is more than or equal to 10,000,000 yuan, then it is right to be input to
After in a certain risk judgment model answered, then the first corresponding conclusion factor exported is for " enterprise product is stagnant
Pin "；If " accounts receivable remaining sum is less than 20,000,000 yuan " and " inventory balance is less than 10,000,000 yuan " is defeated
After entering in corresponding a certain risk judgment model, the first conclusion factor of output is " enterprise's provisional week
Turn problem ".
Fig. 4 is the structural representation of Enterprise Credit Risk Evaluation device embodiment one of the present invention, such as Fig. 4 institute
Show, the present embodiment provide Enterprise Credit Risk Evaluation device include: extraction module 41, output module 42,
Operation module 43 and merging module 44.
Wherein, extraction module 41, for extract intend trusted enterprise objective circumstances investigation report in objective
Factor, the objective factor intending extracting in the objective circumstances investigation report of trusted enterprise is the first objective factor.
Output module 42, for the first objective factor is input in each risk analysis model of pre-stored, defeated
Go out the first analytical factor that each risk analysis model is corresponding.Output module 42, it is objective to be additionally operable to first
Factor, the first analytical factor are input in each risk judgment model of pre-stored, export each risk and sentence
The first conclusion factor that disconnected model is corresponding.Operation module 43, is used for performing operation until operation exports
Conclusion factor no longer change till, operate and be: by the first objective factor, the first analytical factor, upper one
First conclusion factor of secondary output is input in each risk judgment model of pre-stored, exports each risk
The first conclusion factor that judgment models is corresponding.Merge module 44, for by the first objective factor, first point
Plain, whole the first conclusion factor of factorial merges, and forms the credit risk view intending trusted enterprise.
Wherein, each risk analysis model of pre-stored and each risk judgment model of pre-stored are bases
Duo Pian bank responsible investigation report at no distant date builds.
The Enterprise Credit Risk Evaluation device that the present embodiment provides can perform embodiment of the method shown in Fig. 1
Technical scheme, it is similar with technique effect that it realizes principle, and here is omitted.
Fig. 5 is the structural representation of Enterprise Credit Risk Evaluation device embodiment two of the present invention, such as Fig. 5 institute
Show, further, the present embodiment on the basis of Enterprise Credit Risk Evaluation device embodiment one of the present invention,
Also include: cluster module 51 and structure module 52.
Further, extraction module 41, be additionally operable to output module 42 by the first objective factor, first point
Factorial element is input in each risk judgment model of pre-stored, exports each risk judgment model corresponding
Before first conclusion factor, extract the second objective factor in Duo Pian bank responsible investigation report at no distant date,
Second analytical factor and the second conclusion factor.Cluster module 51, for the second objective factor, second point
Factorial element and the second conclusion factor cluster respectively, form objective factor collection, analytical factor collection, conclusion
Set of factors.Build module 52, for building risk analysis decision-making according to objective factor collection and analytical factor collection
Tree-model.Build module 52, be additionally operable to according to objective factor collection, analytical factor collection and conclusion set of factors structure
Build risk judgment decision-tree model.
Further, build module 52, specifically for: with the second whole objective factor collection for input,
The second whole analytical factor collection is output, uses pruning algorithms or rear pruning algorithms in advance to build multiple wind
Danger analysis decision tree-model.
Further, build module 52, specifically for: with whole objective factor collection, whole analyses
Factor and other conclusion set of factors in addition to the conclusion set of factors as output are input, multiple conclusion factors
Collection is output, uses pruning algorithms or rear pruning algorithms in advance, builds multiple risk judgment decision-tree model.
Further, extraction module 41 specifically for: according to the structure of " noun phrase+numeral phrase ",
Extract the second objective factor in Duo Pian bank responsible investigation report at no distant date；According to " noun phrase
+ adjective " structure, extract the second analysis in Duo Pian bank responsible investigation report at no distant date because of
Element；According to the structure of " noun phrase+verb phrase ", extract the responsible tune of Duo Pian bank at no distant date
Look into the second conclusion factor in report.
The Enterprise Credit Risk Evaluation device that the present embodiment provides can perform embodiment of the method shown in Fig. 2
Technical scheme, it is similar with technique effect that it realizes principle, and here is omitted.
One of ordinary skill in the art will appreciate that: realize all or part of step of above-mentioned each method embodiment
Suddenly can be completed by the hardware that programmed instruction is relevant.Aforesaid program can be stored in a computer can
Read in storage medium.This program upon execution, performs to include the step of above-mentioned each method embodiment；And
Aforesaid storage medium includes: ROM, RAM, magnetic disc or CD etc. are various can store program code
Medium.
Last it is noted that various embodiments above is only in order to illustrate technical scheme, rather than right
It limits；Although the present invention being described in detail with reference to foregoing embodiments, this area common
Skilled artisans appreciate that the technical scheme described in foregoing embodiments still can be modified by it,
Or the most some or all of technical characteristic is carried out equivalent；And these amendments or replacement, and
The essence not making appropriate technical solution departs from the scope of various embodiments of the present invention technical scheme.
Claims (10)
1. an Enterprise Credit Risk Evaluation method, it is characterised in that including:
Extract the objective factor in the objective circumstances investigation report intending trusted enterprise, described plan trusted enterprise
The objective factor extracted in objective circumstances investigation report is the first objective factor；
Described first objective factor is input in each risk analysis model of pre-stored, export described often
The first analytical factor that individual risk analysis model is corresponding；
Described first objective factor, described first analytical factor are input to each risk judgment of pre-stored
In model, export the first conclusion factor that described each risk judgment model is corresponding；
Performing operation until the first conclusion factor of described operation output no longer changes, described operation is:
By described first objective factor, described first analytical factor, the first conclusion factor of described last output
It is input in each risk judgment model of described pre-stored, exports described each risk judgment model corresponding
The first conclusion factor；
Described first objective factor, described first analytical factor, the first whole conclusion factors are closed
And, form the credit risk view of described plan trusted enterprise；
Wherein, each risk analysis model of described pre-stored and each risk judgment mould of described pre-stored
Type is to build according to Duo Pian bank responsible investigation report at no distant date.
Method the most according to claim 1, it is characterised in that described by described first objective factor,
Described first analytical factor is input in each risk judgment model of pre-stored, exports described each risk
Before the first conclusion factor that judgment models is corresponding, also include:
Extract the second objective factor in Duo Pian bank responsible investigation report at no distant date, the second analytical factor
With the second conclusion factor；
Described second objective factor, described second analytical factor and described second conclusion factor are carried out respectively
Cluster, forms objective factor collection, analytical factor collection, conclusion set of factors；
Risk analysis decision-tree model is built according to described objective factor collection and described analytical factor collection；
Risk judgment is built according to described objective factor collection, described analytical factor collection and described conclusion set of factors
Decision-tree model.
Method the most according to claim 2, it is characterised in that described according to described objective factor collection
Build risk analysis decision-tree model with described analytical factor collection to specifically include:
With whole described objective factor collection for input, whole described analytical factor collection is output, uses
Pruning algorithms or rear pruning algorithms build multiple risk analysis decision-tree models in advance.
Method the most according to claim 2, it is characterised in that described according to described objective factor,
Described analytical factor collection and described conclusion set of factors build risk judgment decision-tree model and specifically include:
Using whole described objective factor collection, whole described analytical factors and except as export conclusion because of
Other conclusion set of factors outside element collection are input, and multiple described conclusion set of factors are output, use and cut in advance
Branch algorithm or rear pruning algorithms, build multiple risk judgment decision-tree model.
5. according to the method described in any one of claim 2-4, it is characterised in that described extraction is recent
Interior the second objective factor, the second analytical factor and the second conclusion in the responsible investigation report of Duo Pian bank
Factor specifically includes:
According to the structure of " noun phrase+numeral phrase ", extract Duo Pian bank responsible investigation at no distant date
The second objective factor in report；
According to the structure of " noun phrase+adjective ", extract Duo Pian bank responsible investigation at no distant date
The second analytical factor in report；
According to the structure of " noun phrase+verb phrase ", extract the responsible tune of Duo Pian bank at no distant date
Look into the second conclusion factor in report.
6. an Enterprise Credit Risk Evaluation device, it is characterised in that including:
Extraction module, for extracting the objective factor in the objective circumstances investigation report intending trusted enterprise, institute
The objective factor extracted in the objective circumstances investigation report of Shu Ni trusted enterprise is the first objective factor；
Output module, for being input to each risk analysis model of pre-stored by described first objective factor
In, export the first analytical factor that described each risk analysis model is corresponding；
Output module, is additionally operable to be input to prestore by described first objective factor, described first analytical factor
In each risk judgment model of storage, export the first conclusion factor that described each risk judgment model is corresponding；
Operation module, for performing operation until the first conclusion factor of described operation output is no longer changed to
Only, described operation is: by described first objective factor, described first analytical factor, described last time defeated
The the first conclusion factor gone out is input in each risk judgment model of described pre-stored, exports described each
The first conclusion factor that risk judgment model is corresponding；
Merge module, for by described first objective factor, described first analytical factor, whole first
Conclusion factor merges, and forms the credit risk view of described plan trusted enterprise；
Wherein, each risk analysis model of described pre-stored and each risk judgment mould of described pre-stored
Type is to build according to Duo Pian bank responsible investigation report at no distant date.
Device the most according to claim 6, it is characterised in that also include: cluster module and structure
Module；
Described extraction module, be additionally operable to described output module by described first objective factor, described first point
Factorial element is input in each risk judgment model of pre-stored, exports described each risk judgment model pair
Before the first conclusion factor answered, extract second in Duo Pian bank responsible investigation report at no distant date objective
Factor, the second analytical factor and the second conclusion factor；
Described cluster module, for described second objective factor, described second analytical factor and described the
Two conclusion factors cluster respectively, form objective factor collection, analytical factor collection, conclusion set of factors；
Described structure module, divides for building risk according to described objective factor collection and described analytical factor collection
Analysis decision-tree model；
Described structure module, is additionally operable to according to described objective factor collection, described analytical factor collection and described knot
Opinion set of factors builds risk judgment decision-tree model.
Device the most according to claim 7, it is characterised in that described structure module, specifically for:
With whole described second objective factor collection for input, whole described second analytical factor collection is output,
Pruning algorithms or rear pruning algorithms in advance is used to build multiple risk analysis decision-tree models.
Device the most according to claim 7, it is characterised in that described structure module, specifically for:
Using whole described objective factor collection, whole described analytical factors and except as export conclusion set of factors
Other outer conclusion set of factors are input, and multiple described conclusion set of factors are output, use beta pruning in advance to calculate
Method or rear pruning algorithms, build multiple risk judgment decision-tree model.
10. according to the device described in any one of claim 7-9, it is characterised in that described extraction module
Specifically for:
According to the structure of " noun phrase+numeral phrase ", extract Duo Pian bank responsible investigation at no distant date
The second objective factor in report；According to the structure of " noun phrase+adjective ", extract at no distant date
Duo Pian bank responsible investigation report in the second analytical factor；According to " noun phrase+verb phrase "
Structure, extract the second conclusion factor in Duo Pian bank responsible investigation report at no distant date.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610515496.5A CN106022915A (en) | 2016-07-01 | 2016-07-01 | Enterprise credit risk assessment method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610515496.5A CN106022915A (en) | 2016-07-01 | 2016-07-01 | Enterprise credit risk assessment method and apparatus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106022915A true CN106022915A (en) | 2016-10-12 |
Family
ID=57106258
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610515496.5A Pending CN106022915A (en) | 2016-07-01 | 2016-07-01 | Enterprise credit risk assessment method and apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106022915A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108734569A (en) * | 2018-04-17 | 2018-11-02 | 天逸财金科技服务（武汉）有限公司 | A kind of factoring information investigation system and method |
| CN109961198A (en) * | 2017-12-25 | 2019-07-02 | 北京京东尚科信息技术有限公司 | Related information generation method and device |
| CN110110744A (en) * | 2019-03-27 | 2019-08-09 | 平安国际智慧城市科技股份有限公司 | Text matching method, device and computer equipment based on semantic understanding |
-
2016
- 2016-07-01 CN CN201610515496.5A patent/CN106022915A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109961198A (en) * | 2017-12-25 | 2019-07-02 | 北京京东尚科信息技术有限公司 | Related information generation method and device |
| CN109961198B (en) * | 2017-12-25 | 2021-12-31 | 北京京东尚科信息技术有限公司 | Associated information generation method and device |
| CN108734569A (en) * | 2018-04-17 | 2018-11-02 | 天逸财金科技服务（武汉）有限公司 | A kind of factoring information investigation system and method |
| CN110110744A (en) * | 2019-03-27 | 2019-08-09 | 平安国际智慧城市科技股份有限公司 | Text matching method, device and computer equipment based on semantic understanding |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108492001A (en) | A method of being used for guaranteed loan network risk management | |
| Ereiz | Predicting default loans using machine learning (OptiML) | |
| CN111160745A (en) | User account data processing method and device | |
| EP3726441A1 (en) | Company bankruptcy prediction system and operating method therefor | |
| US20190392295A1 (en) | Information processing device, method, and program that use deep learning | |
| KR102628559B1 (en) | Method and device for providing real estate mortgage loan automatic screening platform | |
| CN110930038A (en) | Loan demand identification method, loan demand identification device, loan demand identification terminal and loan demand identification storage medium | |
| CN106022915A (en) | Enterprise credit risk assessment method and apparatus | |
| CN114638696A (en) | Credit risk prediction model training method and system | |
| Kanaparthi | AI-based personalization and trust in digital finance | |
| Khalik Salman et al. | Testing for Granger causality between industrial output and CPI in the presence of regime shift: Swedish data | |
| KR20210029326A (en) | Apparatus and method for diagnosing soundness of company using unstructured financial information | |
| Kothandapani | Applications of Robotic Process Automation in Quantitative Risk Assessment in Financial Institutions | |
| Reddy et al. | Machine Learning based Loan Eligibility Prediction using Random Forest Model | |
| JP6343475B2 (en) | Value chain analysis apparatus and method | |
| CN117114812A (en) | Financial product recommendation method and device for enterprises | |
| CN117290508A (en) | Post-loan text data processing method and system based on natural language processing | |
| KR20110114181A (en) | Loan underwriting method for improving forecasting accuracy | |
| CN117132383A (en) | Credit data processing method, device, equipment and readable storage medium | |
| CN111046947A (en) | Training system and method of classifier and identification method of abnormal sample | |
| Huang et al. | Domain adaptation approach for credit risk analysis | |
| Ahmed et al. | Estimation of causal relationship between world gold prices and KSE 100 index: Evidence from Johansen cointegration technique | |
| TWM583089U (en) | Smart credit risk assessment system | |
| KR102543211B1 (en) | Company's growth potential prediction system using unstructured data | |
| Barjaktarovic et al. | Possibilities of financial support to small and medium hotel companies in Serbie |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161012 |
|
| RJ01 | Rejection of invention patent application after publication |