CN107424026A - Businessman's reputation evaluation method and device - Google Patents

Businessman's reputation evaluation method and device Download PDF

Info

Publication number
CN107424026A
CN107424026A CN201610349194.5A CN201610349194A CN107424026A CN 107424026 A CN107424026 A CN 107424026A CN 201610349194 A CN201610349194 A CN 201610349194A CN 107424026 A CN107424026 A CN 107424026A
Authority
CN
China
Prior art keywords
prestige
credit rating
businessman
achievement data
index item
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
Application number
CN201610349194.5A
Other languages
Chinese (zh)
Inventor
高显
崔燕
杨自强
谭静
于倩
白霄骅
岳爱珍
赵辉
王私江
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Baidu Netcom Science and Technology Co Ltd
Original Assignee
Beijing Baidu Netcom Science and Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Baidu Netcom Science and Technology Co Ltd filed Critical Beijing Baidu Netcom Science and Technology Co Ltd
Priority to CN201610349194.5A priority Critical patent/CN107424026A/en
Publication of CN107424026A publication Critical patent/CN107424026A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0609—Buyer or seller confidence or verification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0641—Shopping interfaces

Abstract

The embodiment of the invention discloses a kind of businessman's reputation evaluation method and device.Methods described includes：Obtain the prestige achievement data of each prestige index item for characterizing businessman's prestige to be evaluated；Acquired prestige achievement data is normalized；According to the credit rating model, the credit rating grade of the prestige achievement data calculating businessman to be evaluated after normalized is utilized；Credit rating grade according to being calculated shows corresponding prestige mark on webpage.Businessman's reputation evaluation method and device provided in an embodiment of the present invention give more structurally sound evaluation to the degrees of comparison of internet businessman.

Description

Businessman's reputation evaluation method and device
Technical field
The present embodiments relate to microcomputer data processing field, more particularly to a kind of businessman's reputation evaluation method and Device.
Background technology
Sincere risk and wooden horse that false identities, dolus malus, fishing website in face of Internet era etc. become increasingly conspicuous, Virus, privacy such as steal at the security threat of diversification, and internet site and ecommerce are faced with the two of " sincerity " and " trust " It is necessary that big challenge, the sincerity for how ensuring website operator and the trust for how obtaining netizen have become internet practitioner The two big pressing issues to be solved.
The Internet portal that search engine is commonly used as most netizens, itself has it in terms of internet data access The advantage that his website hardly matches.Therefore, different internet businessman is carried out by search engine credit rating quietly into For a kind of trend of internet industry.
However, because the crawl to internet data is not comprehensive enough, the many-side such as do not segment to the industry background of businessman The reason for, existing reputation evaluation method is all not accurate enough there is grading unavoidably, it is difficult to the defects of making netizen convince.
The content of the invention
For above-mentioned technical problem, the embodiments of the invention provide the reputation evaluation method of businessman a kind of and device, with right The degrees of comparison of internet businessman provides more structurally sound evaluation.
On the one hand, the embodiments of the invention provide a kind of businessman's reputation evaluation method, methods described to include：
Obtain the prestige achievement data of each prestige index item for characterizing businessman's prestige to be evaluated, each prestige Index item is the input variable of the credit rating model built in advance；
Acquired prestige achievement data is normalized；
According to the credit rating model, calculated using the prestige achievement data after normalized to be evaluated The credit rating grade of businessman；
Credit rating grade according to being calculated shows corresponding prestige mark on webpage.
On the other hand, the embodiment of the present invention additionally provides a kind of businessman's credit rating device, and described device includes：
Acquisition module, for obtaining the prestige index number for each prestige index item for being used to characterize businessman's prestige to be evaluated According to each prestige index item is the input variable of the credit rating model built in advance；
Normalized module, for acquired prestige achievement data to be normalized；
Computing module, for according to the credit rating model, utilizing the prestige index after normalized Data calculate the credit rating grade of businessman to be evaluated；
Display module, for showing corresponding prestige mark on webpage according to the credit rating grade being calculated.
Businessman's reputation evaluation method and device provided in an embodiment of the present invention, it is used to characterize business by excavating from internet The prestige index set of family's prestige, the achievement data in the prestige index set is carried out according to previously given prestige criterion Normalization conversion, according to given credit rating model, the prestige that the achievement data after being changed using normalization calculates businessman is commented Valency grade, and corresponding prestige is showed on webpage according to the credit rating grade being calculated and marked so as to businessman's The evaluation of degrees of comparison is more reliable, credible.
Brief description of the drawings
By reading the detailed description made to non-limiting example made with reference to the following drawings, of the invention is other Feature, objects and advantages will become more apparent upon：
Fig. 1 is the flow chart for businessman's reputation evaluation method that first embodiment of the invention provides；
Fig. 2A is the prestige mark that a kind of embodiment of first embodiment of the invention provides；
Fig. 2 B are the prestige marks that the another embodiment of first embodiment of the invention provides；
Fig. 2 C are the prestige marks that the another embodiment of first embodiment of the invention provides；
Fig. 3 is the flow chart of normalized in businessman's reputation evaluation method that second embodiment of the invention provides；
Fig. 4 is the flow chart for businessman's reputation evaluation method that third embodiment of the invention provides；
Fig. 5 is the flow chart for businessman's reputation evaluation method that fourth embodiment of the invention provides；
Fig. 6 is the flow chart that operation is obtained in businessman's reputation evaluation method that fifth embodiment of the invention provides；
Fig. 7 is the flow chart for businessman's reputation evaluation method that sixth embodiment of the invention provides；
Fig. 8 is the structure chart for businessman's credit rating device that seventh embodiment of the invention provides.
Embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched The specific embodiment stated is used only for explaining the present invention, rather than limitation of the invention.It also should be noted that in order to just Part related to the present invention rather than entire infrastructure are illustrate only in description, accompanying drawing.
First embodiment
Present embodiments provide a kind of technical scheme of businessman's reputation evaluation method.Businessman's prestige that the present embodiment provides is commented Valency method is performed by businessman's credit rating device.Businessman's credit rating device is generally integrated in a server of network side In.
Referring to Fig. 1, businessman's reputation evaluation method includes：
S11, the prestige achievement data of each prestige index item for characterizing businessman's prestige to be evaluated is obtained, it is described each Prestige index item is the input variable of the credit rating model built in advance.
The prestige achievement data is one group of data target for being used to characterize the prestige of businessman.It is selected into as the prestige The data target of achievement data, it should be that can represent the horizontal data of the prestige of a businessman in industry to a certain extent to refer to Mark.
When carrying out businessman's credit rating to businessman, it should evaluated according to a credit rating model.And the prestige The each prestige index item included in achievement data is then the input variable of the credit rating model, that is, is used to carry out businessman's letter Praise the evaluation variable of evaluation.In one implementation, for all businessmans, unified credit rating mould can be built in advance Type.
It is understood that by the excavation to internet data, can get specific in above-mentioned prestige achievement data The value of prestige index item.For example, it can obtain above-mentioned prestige by capturing, analyzing the text data on businessman website and refer to Mark the value of the specific prestige index item in data.For example, on the website of a businessman oneself, having such text, " I takes charge of The existing people of headcount 392 ".By the analysis of the text to this, the existing headcount of the businessman can be learnt.
In addition, by the access of the database to locally having been built up, it can get in above-mentioned prestige achievement data and have The value of body prestige index item.For example, the businessman feature database that can be established by access needle to different businessmans, described in acquisition The value of specific prestige index item in prestige achievement data.And the businessman feature database can then be submitted by businessman oneself Data establish.
S12, acquired prestige achievement data is normalized.
Because the original index data excavated first typically belong to unstructured data, so in the prestige index set Original achievement data representation is more random under normal circumstances, without the unified form of expression.
For example, the profit excavated to different internet logistics businessmans, some are likely to be using member as measurement unit, And some is then likely to be with ten thousand yuan as measurement unit.For such situation, it is necessary to enter to different data representations Row is unified, that is, normalization conversion.
Still further aspect, because the credit rating model of use can be to the prestige index item in the prestige achievement data Value is weighted average, and the result of weighted average in order to make different businessmans final has comparability, by each information Index item is substituted into before weighted average formula, allows the value of different prestige index item to have unified codomain as far as possible.It is this to not With the unification of the value codomain of prestige index item, and another content of normalization conversion.
After operation by above-mentioned normalization conversion, the achievement data in the prestige index set has unified performance Form, unified value codomain, this is beneficial to carry out credit rating according to unified standard to different businessmans.
S13, according to the credit rating model, calculated and treated using the prestige achievement data after normalized Evaluate the credit rating grade of businessman.
Specifically, the credit rating model is Logistic regression models.By the index number after the normalization conversion According to the previously given Logistic regression models of substitution, it becomes possible to obtain the credit rating grade of businessman.
S14, corresponding prestige is showed on webpage according to the credit rating grade being calculated and marked.
Different credit rating grades corresponds to different prestige and marked.For example, the prestige mark can be highest letter Praise prestige mark corresponding to opinion rating.Or the prestige mark can also be prestige corresponding to time high credit rating grade Mark.
Moreover, show the prestige mark on webpage, can be as shown in Fig. 2A, with the shape of Natural check 21 Formula shows；It can also be showed as Fig. 2 B are shown in the form of floating layer 22；Can also be as Fig. 2 C be shown, with letter The form for praising the object information label 23 of platform is showed.
The present embodiment is right by obtaining the prestige achievement data for being used for characterizing each prestige index item for evaluating businessman's prestige Acquired prestige achievement data is normalized, according to the credit rating model, using described by normalization Prestige achievement data after reason calculates the credit rating grade of businessman to be evaluated, and according to the credit rating grade being calculated Show corresponding prestige mark on webpage so that the evaluation to the degrees of comparison of businessman is more reliable, credible.
Second embodiment
The present embodiment further provides the businessman credit rating side based on the above-mentioned first embodiment of the present invention A kind of technical scheme of normalized in method.In the technical scheme, acquired prestige achievement data is normalized Conversion includes：Statistical analysis is carried out to acquired prestige achievement data, to obtain the statistical nature of each prestige achievement data； Based on the statistical nature of resulting each prestige achievement data, it is determined that corresponding normalized mode；Based on identified Normalized mode, corresponding prestige achievement data is normalized.
Referring to Fig. 3, conversion, which is normalized, to acquired prestige achievement data includes：
S131, statistical analysis is carried out to acquired prestige achievement data, to obtain the statistics of each prestige achievement data Feature.
It is understood that prestige index item different in the prestige achievement data, can have different statistical natures. For example, the statistical nature of rayleigh distributed can be presented in this prestige index item of the moon sales volume of internet retailer.And internet The statistical nature of normal distribution can be presented in this prestige index item of the moon profit of loglstics enterprise.Because different statistical natures is rear Correspond to different normalized modes in continuous normalized, therefore, when really performing above-mentioned normalized, need Above-mentioned statistical nature is handled.
Identification to above-mentioned statistical nature can be according to the statistical analysis to the identical prestige index item of different businessmans.Moreover, In above-mentioned statistical analysis, the data volume of collection is bigger, then the identification for statistical nature is more accurate.
S132, based on the statistical nature of resulting each prestige achievement data, it is determined that corresponding normalized mode.
In normalized, it is necessary to by the original value of different prestige index item map to one it is fixed, in advance In the value codomain of definition.And the prestige index item with different statistical natures is same as, above-mentioned mapping mode is different, that is, The processing mode of normalized is different.For example, for the prestige index item in normal distribution, the normalizing of normal distribution is corresponding with Change processing mode, and for the prestige index item in rayleigh distributed, it is corresponding with the prestige index item of rayleigh distributed.
In addition to above-mentioned value mapping, normalized may also include the behaviour of some data representations conversion Make.For example, for this prestige index item of the profit of loglstics enterprise, this first measurement unit has been employed.And to one In the mining process of the prestige index of specific loglstics enterprise, actual excavation to data used by measurement unit be ten thousand yuan.It is right Just need to perform form of expression conversion in such data.
For above-mentioned form of expression conversion operation, in the step of normalized mode determines, need also exist for into one Step is clear and definite.
More specifically, when determining the step of needing to perform in normalized, the operation that each step needs, and respectively , can be by the action type of each step after parameter corresponding to individual operation, the parameter needed for each operation is with the shape of chained list Formula is stored, and basic data is provided for the normalized of subsequent execution.
S133, based on identified normalized mode, corresponding prestige achievement data is normalized.
After the normalized mode of specific prestige index item is determined, according to fixed normalized side Formula, the value of above-mentioned prestige index item is normalized.
Specifically, when performing the normalized, it can read and determine to deposit in step in foregoing normalized mode The chained list of storage, the action type of each step in normalized, and the parameter needed for each operation are therefrom obtained, so it is complete Into the normalized.
The present embodiment has obtained each prestige achievement data by carrying out statistical analysis to acquired prestige achievement data Statistical nature, based on the statistical nature of resulting each prestige achievement data, it is determined that corresponding normalized mode, with And based on identified normalized mode, corresponding prestige achievement data is normalized so that substitute into prestige Prestige index item in evaluation model has unified data mode, avoids due to being commented caused by the disunity on data mode Valency mistake.
3rd embodiment
The present embodiment further provides businessman's prestige based on the above-mentioned first or second embodiments of the present invention A kind of technical scheme of evaluation method.In the technical scheme, the credit rating model can also include index item weight because Son and/or the statistical nature factor, the index item weight factor are used to characterize the index item when carrying out businessman's credit rating Importance, the statistical nature factor is used to characterize the influence of the statistical nature of the index item to businessman's credit rating.Phase Ying Di, according to the credit rating model, calculated using the prestige achievement data after normalized to be evaluated Before the credit rating grade of businessman, businessman's reputation evaluation method also includes：Each index is determined based on analytic hierarchy process (AHP) The index item weight factor of item；And/or the statistical nature based on each index item determines the statistical nature factor of each index item.
Fig. 4 shows the flow chart of the prioritization scheme based on first embodiment.Referring to Fig. 4, businessman's prestige is commented Valency method includes：
S41, the finger of each index item is determined based on analytic hierarchy process (AHP) (Analytic hierarchy process, AHP) Mark item weight factor.
In the present embodiment, before using the credit rating model, it is necessary first to determine the credit rating model In index item weight factor and/or the statistical nature factor.
The credit rating model is Logistic regression models.Specifically, the Logistic regression models can be Conventional Logistic regression models or follow-on Logistic regression models.No matter which kind of Logistic is used Regression model, before the output valve using sigmoid function computation models, it is required for each input variable, that is, institute Prestige index item is stated to be weighted averagely.
For the Logistic regression models of routine, it is conventional prestige index item to input prestige index item therein, corresponding Weighted factor also only include index item weight factor.Its average weighted calculation formula is given by：
Z=w0f(x0)+w1f(x1)+Λ+wNf(xN)
Wherein, xiIt is i-th of conventional prestige index item in the prestige achievement data, f (xi) it is i-th of conventional prestige Mapping function value corresponding to index item, wiIt is then the index item weight factor corresponding to the conventional prestige index item.And z be then In the Logistic regression models, the input value of sigmoid functions.
For follow-on Logistic regression models, prestige index item therein is inputted just no longer only comprising conventional letter Index item is praised, will also include the prestige index item with statistical nature.And these prestige index item with statistical nature plus Weight factor, not only including index item weight factor, in addition to the statistical nature factor.Follow-on Logistic regression models Weighted average calculation formula provided by equation below：
Z=v0w0f(y0)+v1w1f(y1)+Λ+vMwMf(yM)
Wherein, yjBe j-th have statistical nature prestige index item, f (yj) it is that this has the prestige index of statistical nature Function Mapping value corresponding to, wjIt is the index item weight factor corresponding to j-th of prestige index item, vjIt is that this has statistics special The statistical nature factor corresponding to the prestige index item of sign.And z is then the sigmoid functions in the Logistic regression models Input value.Wherein, vjValue be more than 0, less than or equal to 1.
In above-mentioned follow-on Logistic regression models, the index item weight factor and the statistical nature because Son needs to meet following condition：
v0w0+v1w1+Λ+vMwM=1
That is, either using conventional Logistic regression models, or using modified Logistic recurrence Model, it is required to obtain the index item weight factor in model.And only when using modified Logistic regression models, just need Obtain the statistical nature factor in model.
More specifically, the acquisition to index item weight factor is according to AHP modes.AHP resolves into challenge each Compositing factor, and these factor case dominance relations are grouped to form recursive hierarchy structure.Layer is determined by way of comparing two-by-two The relative importance of secondary middle factors.Then the judgement of integrated decision-making person, total sequence of decision scheme relative importance is determined.
Moreover, the businessman for belonging to different industries type, obtains its corresponding index using independent AHP processes and weighs Repeated factor.So, for the businessman of different industries type, the prestige index item in its credit rating model would be possible to not Together.For example, to internet loglstics enterprise, its prestige index item includes headcount, and for IC design enterprise, its Do not include headcount in prestige index item.
S42, acquired prestige achievement data is normalized.
S43, the statistical nature based on each index item determine the statistical nature factor of each index item.
The step is suitable only for follow-on Logistic regression models, for returning mould using conventional Logistic The situation of type, it is not necessary to perform the operation.
The prestige index item with statistical nature is so a kind of index item, and the index item is described follow-on The value of the statistical nature factor in Logistic regression models has close relationship with the statistical nature of the index item.
For example, for the index item in normal distribution, the value of the distribution statistical nature factor is 0.2.And in Rayleigh The index item of distribution, the value of the distribution statistical nature factor is 0.1.
S44, according to the credit rating model, calculated and treated using the prestige achievement data after normalized Evaluate the credit rating grade of businessman.
S45, the prestige achievement data of each prestige index item for characterizing businessman's prestige to be evaluated is obtained, it is described each Prestige index item is the input variable of the credit rating model built in advance.
S46, corresponding prestige is showed on webpage according to the credit rating grade being calculated and marked.
The present embodiment is by the way that before prestige achievement data is obtained, the index of each index item is determined based on analytic hierarchy process (AHP) Item weight factor, and the statistical nature based on each index item determine the statistical nature factor of each index item so that are obtaining Before taking credit rating model described in input value, two important parameters of the credit rating model are determined.
Fourth embodiment
The present embodiment is a kind of optimisation technique scheme based on the present invention above-mentioned first to 3rd embodiment.Fig. 5 shows A kind of flow chart of the implementation optimized based on first embodiment is gone out.
Referring to Fig. 5, businessman's reputation evaluation method includes：
S51, the prestige achievement data of each prestige index item for characterizing businessman's prestige to be evaluated is obtained, it is described each Prestige index item is the input variable of the credit rating model built in advance.
S52, data cleansing is carried out to acquired prestige achievement data.
The purpose of data cleansing retains available finger in the achievement data that falseness is rejected from the achievement data grabbed Mark data.For example, in achievement data comprising whether the achievement data of state-owned enterprise.And the value of this index be only possible to be it is yes, Or it is no.If other values outside said two devices occurs in the value of this achievement data, this index can be concluded that Attribute is falseness, it is necessary to be rejected by data cleansing from Criterion Attribute.
S53, structuring processing and integration are carried out to acquired prestige achievement data.
Structuring processing refers to the form of expression of non-structured data carrying out unification.For example, from it is non-structured mutually The prestige achievement data of needs, and the letter arrived with given data structure storage said extracted are extracted in networking natural language text Praise achievement data.
Integration processing refers to be integrated discrete data division.For example, one internet logistics businessman's of excavation During area data of storing in a warehouse, the East storage area for obtaining the businessman is 100,000 square metres, and West storage area is 200,000 square metres. Now, it should be 300,000 square metres in the storage gross area that integration processing is the businessman.
S54, the data after integration are verified.
In order to ensure the achievement data in the prestige index set excavated accurately and reliably, it is necessary to the index after integration processing Data are verified.
The verification can be according to the verification of this index data characteristics of itself, for example, the data of a certain index take It is worth unit between 1 to 100, then the achievement data can be verified according to this feature of this achievement data.Institute State the verification that verifies and can also be according to reference data.For example, there is preset reference data for the achievement data of some businessman Backup, then the achievement data excavated can be verified according to above-mentioned reference data backup.
S55, based on the source of the prestige achievement data, differentiation renewal is carried out to the prestige achievement data obtained.
Because the indices of businessman are continually changing with the passage of time, so being excavated from different data sources , it is necessary to according to the change of derived data after to the primary data of prestige index set, timely achievement data is updated.
So that derived data is internet web page as an example, if producing the initial data hair of the achievement data in source page Change is given birth to, then the achievement data should also change therewith.
Specifically, periodically the data on primary source webpage can be checked using crawlers.It is once it was found that former Data on beginning source page are updated, then the achievement data are excavated again, to ensure achievement data more Newly.
Currently, in addition to internet web page, the derived data can also have other forms.Such as the source number According to can be according to businessman submit data establish local data.So-called differentiation renewal refers to, for different derived datas Property, using the different update cycles.For example, the renewal frequency of internet web page is typically very fast, so to internet web page Using the renewal granularity less update cycle.And the renewal frequency of local data base is slower, then to local data base using renewal The granularity larger update cycle.
S56, acquired prestige achievement data is normalized.
S57, according to the credit rating model, calculated and treated using the prestige achievement data after normalized Evaluate the credit rating grade of businessman.
S58, corresponding prestige is showed on webpage according to the credit rating grade being calculated and marked.
The present embodiment to acquired prestige achievement data by carrying out data cleansing, to acquired prestige achievement data Structuring processing and integration are carried out, the data after integration are verified, based on the source of the prestige achievement data, to being obtained Prestige achievement data carry out differentiation renewal, ensure that the accuracy of the prestige achievement data got and ageing.
, can also be to businessman's credit rating for being shown in Fig. 5 it will be clear that in the other manner of the present embodiment Scheme carries out following various modifications：(1) delete step S52 to S55；(2) delete step S55；(3) delete step S51 and S55； (4) delete step S52 to S54；(5) delete step S53 and S54.
5th embodiment
The present embodiment is a kind of optimisation technique scheme based on the present invention above-mentioned first to fourth embodiment.Specifically Ground, in the technical scheme, the credit rating model includes one or more credit rating models corresponding with industry type. For the difference of industry field where the businessman, the credit rating model actually used is different, also, further described Contain different prestige index item in prestige achievement data.For example, for internet house lease businessman, the prestige Index item should include the source of houses quantity that this businessman possesses.And for internet logistics businessman, the prestige index item The logistic storage capacity of this businessman should be included.
Correspondingly, the prestige achievement data for obtaining the prestige index item for being used for characterizing businessman's prestige to be evaluated is specifically limited For：The industry type submitted according to businessman, it is determined that corresponding credit rating model；Obtain in determined credit rating model Each prestige index item prestige achievement data.
Referring to Fig. 6, obtaining the prestige achievement data of the prestige index item for characterizing businessman's prestige to be evaluated includes：
S61, the industry type submitted according to businessman, it is determined that corresponding credit rating model.
Businessman is when submitting the basic data of itself, it is necessary to submit the industry type belonging to itself.Further, it is described Industry type is classified according to the general trade mark of our times, namely Nice is classified to determine.
S62, obtain the prestige achievement data of each prestige index item in determined credit rating model.
After the credit rating model determines, obtaining needs input to each prestige in the credit rating model to refer to Mark item.Each prestige index item can be obtained by the excavation to internet web page text, can also be by for example local number Obtained according to the access of other databases in storehouse.
The present embodiment is by the industry type submitted according to businessman, it is determined that corresponding credit rating model, and obtain institute The prestige achievement data of each prestige index item in the credit rating model determined, is realized for belonging to different industries Businessman prestige achievement data acquisition.
Sixth embodiment
A kind of optimisation technique scheme of the present embodiment based on the above-mentioned first to the 5th embodiment of the invention.In the technology In scheme, after credit rating grade is calculated, businessman's reputation evaluation method can also include：Utilize predetermined assessment mould Type, the accuracy of the credit rating grade to being calculated are assessed.
Fig. 7 shows the flow chart of the embodiment optimized based on first embodiment.Referring to Fig. 7, the business Family's reputation evaluation method includes：
S71, the prestige achievement data of each prestige index item for characterizing businessman's prestige to be evaluated is obtained, it is described each Prestige index item is the input variable of the credit rating model built in advance.
S72, acquired prestige achievement data is normalized.
S73, according to the credit rating model, calculated and treated using the prestige achievement data after normalized Evaluate the credit rating grade of businessman.
S74, using predetermined assessment models, the accuracy of the credit rating grade to being calculated is assessed.
Specifically, can be according to standardization accumulation of discount increment (Normalized discounted cumulative Gain, NDCG), the evaluation method such as sequence (Mean reciprocal rank, MRR) reciprocal is to the side that is provided according to the present embodiment The accuracy of the credit rating grade for the businessman that method obtains is assessed.
S75, corresponding prestige is showed on webpage according to the credit rating grade being calculated and marked.
The present embodiment using the prestige achievement data after normalized by calculating businessman's to be evaluated After credit rating grading, using predetermined assessment models, the accuracy of the credit rating grade to being calculated is assessed, So that the prestige mark finally showed has higher accuracy.
7th embodiment
Present embodiments provide a kind of structure chart of businessman's credit rating device.Referring to Fig. 8, businessman's credit rating dress Put including：Acquisition module 802, normalized module 807, computing module 809 and display module 811.
The prestige that the acquisition module 802 is used to obtain each prestige index item for being used for characterizing businessman's prestige to be evaluated refers to Data are marked, each prestige index item is the input variable of the credit rating model built in advance.
The normalized module 807 is used to acquired prestige achievement data be normalized.
The computing module 809 is used for according to the credit rating model, utilizes the letter after normalized Praise the credit rating grade that achievement data calculates businessman to be evaluated.
The display module 811 is used to show corresponding prestige mark on webpage according to the credit rating grade being calculated Note.
Optionally, the normalized module 807 includes：Statistic unit, normalized mode determining unit and return One changes processing unit.
The statistic unit is used to carry out statistical analysis to acquired prestige achievement data, to obtain each prestige index The statistical nature of data.
The normalized mode determining unit is used for the statistical nature based on resulting each prestige achievement data, It is determined that corresponding normalized mode.
The normalized unit is used for the normalized mode based on determined by, to corresponding prestige achievement data It is normalized.
Optionally, for each index item, the credit rating model includes index item weight factor, the index item power Repeated factor is used to characterize importance of the index item when carrying out businessman's credit rating, and described device also includes：Weight Factor determining module 801.
The weight factor determining module 801 be used for based on analytic hierarchy process (AHP) determine the index item weight of each index item because Son.
Optionally, for each index item, the credit rating model includes the statistical nature factor, the statistical nature because Influence of the statistical nature that son is used to characterize the index item to businessman's credit rating, and described device also include：Statistics is special Levy factor determining module 808.
The statistical nature factor determining module 808 determines each index item for the statistical nature based on each index item The statistical nature factor.
Optionally, the credit rating model includes one or more credit rating models corresponding with industry type, with And the acquisition module 802 includes：Credit rating model determining unit and prestige achievement data acquiring unit.
The credit rating model determining unit is used for the industry type submitted according to businessman, it is determined that corresponding credit rating Model.
Each prestige that the prestige achievement data acquiring unit is used to obtain in determined credit rating model refers to Mark the prestige achievement data of item.
Optionally, businessman's credit rating device also includes：Data cleansing module 803.
The data cleansing module 803 is used to carry out data cleansing to acquired prestige achievement data.
Optionally, businessman's credit rating device also includes：Integrate module 804.
The module 804 of integrating is used to carry out structuring processing and integration to acquired prestige achievement data.
Optionally, businessman's credit rating device also includes：Correction verification module 805.
The correction verification module 805 is used to verify the data after integration.
Optionally, the credit rating model is Logic Regression Models or modified Logic Regression Models.
Optionally, businessman's credit rating device also includes：Update module 806.
The update module 806 is used to obtain the prestige index for the prestige index item for being used to characterize businessman's prestige to be evaluated After data, based on the source of the prestige achievement data, differentiation renewal is carried out to the prestige achievement data obtained.
Optionally, businessman's credit rating device also includes：Evaluation module 810.
The evaluation module 810 is used to utilize predetermined assessment models, to the accurate of the credit rating grade that is calculated Property is assessed.
Will be appreciated by those skilled in the art that above-mentioned each module of the invention or each step can use general meter Device is calculated to realize, they can be concentrated on single computing device, or are distributed in the network that multiple computing devices are formed On, alternatively, they can be realized with the program code that computer installation can perform, so as to be stored in storage Performed in device by computing device, they are either fabricated to each integrated circuit modules respectively or will be more in them Individual module or step are fabricated to single integrated circuit module to realize.So, the present invention be not restricted to any specific hardware and The combination of software.
Each embodiment in this specification is described by the way of progressive, what each embodiment stressed be with The difference of other embodiment, same or analogous part between each embodiment mutually referring to.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention, for those skilled in the art For, the present invention can have various changes and change.All any modifications made within spirit and principles of the present invention, it is equal Replace, improve etc., it should be included in the scope of the protection.

Claims (18)

  1. A kind of 1. businessman's reputation evaluation method, it is characterised in that including：
    Obtain the prestige achievement data of each prestige index item for characterizing businessman's prestige to be evaluated, each prestige index Item is the input variable of the credit rating model built in advance；
    Acquired prestige achievement data is normalized；
    According to the credit rating model, businessman to be evaluated is calculated using the prestige achievement data after normalized Credit rating grade；
    Credit rating grade according to being calculated shows corresponding prestige mark on webpage.
  2. 2. according to the method for claim 1, it is characterised in that conversion is normalized to acquired prestige achievement data Including：
    Statistical analysis is carried out to acquired prestige achievement data, to obtain the statistical nature of each prestige achievement data；
    Based on the statistical nature of resulting each prestige achievement data, it is determined that corresponding normalized mode；
    Based on identified normalized mode, corresponding prestige achievement data is normalized.
  3. 3. method according to claim 1 or 2, it is characterised in that for each index item, the credit rating model bag Index item weight factor is included, the index item weight factor is used to characterize weight of the index item when carrying out businessman's credit rating The property wanted, and methods described also include：
    The index item weight factor is determined based on analytic hierarchy process (AHP).
  4. 4. according to the method for claim 3, it is characterised in that for each index item, the credit rating model includes The statistical nature factor, the statistical nature factor are used for shadow of the statistical nature to businessman's credit rating for characterizing the index item Ring, and methods described also includes：
    Statistical nature based on the index item determines the statistical nature factor.
  5. 5. according to the method described in any claim in claim 1-4, it is characterised in that the credit rating model includes One or more credit rating models corresponding with industry type, and obtain and refer to for characterizing the prestige of businessman's prestige to be evaluated The prestige achievement data of mark item includes：
    The industry type submitted according to businessman, it is determined that corresponding credit rating model；
    Obtain the prestige achievement data of each prestige index item in determined credit rating model.
  6. 6. according to the method described in any claim in claim 1-5, it is characterised in that methods described also includes：
    Data cleansing is carried out to acquired prestige achievement data.
  7. 7. according to the method described in any claim in claim 1-6, it is characterised in that methods described also includes：
    Structuring processing and integration are carried out to acquired prestige achievement data.
  8. 8. according to the method for claim 7, it is characterised in that methods described also includes：
    Data after integration are verified.
  9. 9. according to the method described in any claim in claim 1-8, it is characterised in that the credit rating model is to patrol Collect regression model or modified Logic Regression Models.
  10. 10. according to the method described in any claim in claim 1-9, it is characterised in that to be evaluated for characterizing obtaining After the prestige achievement data of the prestige index item of valency businessman's prestige, methods described also includes：
    Based on the source of the prestige achievement data, differentiation renewal is carried out to the prestige achievement data obtained.
  11. 11. according to the method described in any claim in claim 1-10, it is characterised in that methods described also includes：
    Using predetermined assessment models, the accuracy of the credit rating grade to being calculated is assessed.
  12. A kind of 12. businessman's credit rating device, it is characterised in that including：
    Acquisition module, for obtaining the prestige achievement data for each prestige index item for being used to characterize businessman's prestige to be evaluated, institute State the input variable that each prestige index item is the credit rating model built in advance；
    Normalized module, for acquired prestige achievement data to be normalized；
    Computing module, for according to the credit rating model, utilizing the prestige achievement data after normalized Calculate the credit rating grade of businessman to be evaluated；
    Display module, for showing corresponding prestige mark on webpage according to the credit rating grade being calculated.
  13. 13. businessman's credit rating device according to claim 12, it is characterised in that the normalized module bag Include：
    Statistic unit, for carrying out statistical analysis to acquired prestige achievement data, to obtain each prestige achievement data Statistical nature；
    Normalized mode determining unit, for the statistical nature based on resulting each prestige achievement data, it is determined that pair The normalized mode answered；
    Normalized unit, for based on identified normalized mode, returning to corresponding prestige achievement data One change is handled.
  14. 14. businessman's credit rating device according to claim 12 or 13, it is characterised in that for each index item, institute Stating credit rating model includes index item weight factor, and the index item weight factor is being entered to do business for characterizing the index item Importance during family's credit rating, and described device also include：
    Weight factor determining module, for determining the index item weight factor of each index item based on analytic hierarchy process (AHP).
  15. 15. businessman's credit rating device according to claim 14, it is characterised in that for each index item, the letter Reputation evaluation model includes the statistical nature factor, and the statistical nature that the statistical nature factor is used to characterize the index item is to businessman The influence of credit rating, and described device also include：
    Statistical nature factor determining module, the statistical nature of each index item is determined for the statistical nature based on each index item The factor.
  16. 16. businessman's credit rating device according to any claim in claim 12-15, it is characterised in that described Credit rating model includes one or more credit rating models corresponding with industry type, and the acquisition module includes：
    Credit rating model determining unit, for the industry type submitted according to businessman, it is determined that corresponding credit rating model；
    Prestige achievement data acquiring unit, for obtaining the letter of each prestige index item in determined credit rating model Praise achievement data.
  17. 17. businessman's credit rating device according to any claim in claim 12-16, it is characterised in that described Device also includes：
    Update module, after in acquisition for the prestige achievement data for the prestige index item for characterizing businessman's prestige to be evaluated, Based on the source of the prestige achievement data, differentiation renewal is carried out to the prestige achievement data obtained.
  18. 18. businessman's credit rating device according to any claim in claim 12-17, it is characterised in that described Device also includes：
    Evaluation module, for being assessed using the accuracy of predetermined assessment models, the credit rating grade to being calculated.
CN201610349194.5A 2016-05-24 2016-05-24 Businessman's reputation evaluation method and device Pending CN107424026A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610349194.5A CN107424026A (en) 2016-05-24 2016-05-24 Businessman's reputation evaluation method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610349194.5A CN107424026A (en) 2016-05-24 2016-05-24 Businessman's reputation evaluation method and device

Publications (1)

Publication Number Publication Date
CN107424026A true CN107424026A (en) 2017-12-01

Family

ID=60422745

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610349194.5A Pending CN107424026A (en) 2016-05-24 2016-05-24 Businessman's reputation evaluation method and device

Country Status (1)

Country Link
CN (1) CN107424026A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109102206A (en) * 2018-08-31 2018-12-28 深圳市轱辘汽车维修技术有限公司 A kind of evaluation method and relevant device of Automobile Service Factory
CN110766416A (en) * 2018-07-24 2020-02-07 北京京东尚科信息技术有限公司 Method, apparatus, computer system, and medium for merchant ranking
CN111784456A (en) * 2020-06-30 2020-10-16 广东奥园奥买家电子商务有限公司 Reputation evaluation method and device for e-commerce merchant and computer equipment

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102163310A (en) * 2010-02-22 2011-08-24 深圳市腾讯计算机系统有限公司 Information pushing method and device based on credit rating of user

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102163310A (en) * 2010-02-22 2011-08-24 深圳市腾讯计算机系统有限公司 Information pushing method and device based on credit rating of user

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
茹永梅: "基于模糊理论的O2O模式商家信誉评估模型", 《西安邮电大学学报》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110766416A (en) * 2018-07-24 2020-02-07 北京京东尚科信息技术有限公司 Method, apparatus, computer system, and medium for merchant ranking
CN109102206A (en) * 2018-08-31 2018-12-28 深圳市轱辘汽车维修技术有限公司 A kind of evaluation method and relevant device of Automobile Service Factory
CN111784456A (en) * 2020-06-30 2020-10-16 广东奥园奥买家电子商务有限公司 Reputation evaluation method and device for e-commerce merchant and computer equipment

Similar Documents

Publication Publication Date Title
Gondia et al. Machine learning algorithms for construction projects delay risk prediction
Feizizadeh et al. Spatially explicit sensitivity and uncertainty analysis for multicriteria-based vulnerability assessment
US20220374746A1 (en) Model interpretation
Chatterjee et al. Supplier selection in Telecom supply chain management: a Fuzzy-Rasch based COPRAS-G method
Owolabi et al. Predicting completion risk in PPP projects using big data analytics
CN111309822B (en) User identity recognition method and device
CN107230108A (en) The processing method and processing device of business datum
Wang et al. Predicting ENR construction cost index using machine-learning algorithms
Wu et al. Simulating mixed land-use change under multi-label concept by integrating a convolutional neural network and cellular automata: A case study of Huizhou, China
CN110458324B (en) Method and device for calculating risk probability and computer equipment
US11922283B2 (en) Model interpretation
CN112116184A (en) Factory risk estimation using historical inspection data
Cano et al. International market selection using fuzzy weighing and Monte Carlo simulation
US9858526B2 (en) Method and system using association rules to form custom lists of cookies
CN116402512B (en) Account security check management method based on artificial intelligence
CN108876458A (en) A kind of commodity dynamic prediction method and system based on web crawlers
CN106970986A (en) Urban waterlogging influence degree method for digging and system based on deep learning
CN107424026A (en) Businessman's reputation evaluation method and device
CN116401379A (en) Financial product data pushing method, device, equipment and storage medium
CN114117134A (en) Abnormal feature detection method, device, equipment and computer readable medium
CN110310012A (en) Data analysing method, device, equipment and computer readable storage medium
WO2019204072A1 (en) Model interpretation
Feng Data analysis and prediction modeling based on deep learning in E-commerce
Feizizadeh et al. Sensitivity and uncertainty analysis approach for GIS-MCDA based economic vulnerability assessment
CN114219287A (en) Taxpayer risk evaluation method based on graph neural network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20171201

RJ01 Rejection of invention patent application after publication