CN107403326A - A kind of Insurance Fraud recognition methods and device based on teledata - Google Patents
A kind of Insurance Fraud recognition methods and device based on teledata Download PDFInfo
- Publication number
- CN107403326A CN107403326A CN201710694259.4A CN201710694259A CN107403326A CN 107403326 A CN107403326 A CN 107403326A CN 201710694259 A CN201710694259 A CN 201710694259A CN 107403326 A CN107403326 A CN 107403326A
- Authority
- CN
- China
- Prior art keywords
- case
- personnel
- fraud
- insurance
- personnel concerning
- 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
- G06Q30/00—Commerce
- G06Q30/018—Certifying business or products
- G06Q30/0185—Product, service or business identity fraud
-
- 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
- G06Q40/08—Insurance
Abstract
The present invention discloses a kind of Insurance Fraud recognition methods based on teledata, including：Extracted from personnel concerning the case's teledata of insurance case to be identified the personnel concerning the case between any two connect data each other；According to the personnel concerning the case between any two connect data each other, using the first data model being pre-configured with, calculate the coefficient correlation of the personnel concerning the case between any two；According to the coefficient correlation, judge the personnel concerning the case whether close relation；According to the personnel concerning the case whether close relation, judge the insurance case to be identified whether be fraud case.Correspondingly, the invention also discloses a kind of Insurance Fraud identification device based on teledata.The embodiment of the present invention can quick and precisely identify Insurance Fraud, avoid insurance company an innocent person from suffering a loss, and improve the efficiency of claim processing.
Description
Technical field
The present invention relates to fraud identification field, in particular it relates to a kind of Insurance Fraud recognition methods based on teledata
And device.
Background technology
In recent years, Domestic Insurance company vehicle insurance business premium total income rises year by year, but earning performance causes anxiety.Insurance Regulatory Commission unites
Meter shows that the case that 20% is there are about in car insurance claim belongs to fraud.Why auto insurance operation is constantly in loss edge, and one very
The reason for important, is that spreading unchecked for vehicle insurance fraud.Fraud is called the noiseless catastrophe of insurance, is not swallowing all the time
The meagre profit space of insurance, insurance company is set to be lost by an innocent person.Therefore, the implementation of anti-fake system is for promoting to protect
The sound development of dangerous industry and lifting insurance company profitability are all of great importance.
In the traditional process of insurance company's processing claim, using manually to the anti-fraud investigation of each claim case progress；And
In recent years, the anti-fake system of Intelligent Recognition fraud case is proposed both at home and abroad, can be taken advantage of with the height in the numerous claims of Intelligent Recognition
The case of risk is cheated, is investigated further for the case of high risk of fraud.But its data mainly for insurance case in itself,
It is identified using single supervised learning algorithm, regression analysis or other statistical methods, the identification for cheating case is misjudged,
Situation of failing to judge is more.
The content of the invention
The embodiment of the present invention provides a kind of Insurance Fraud recognition methods and device based on teledata, can be quick and precisely
Insurance Fraud is identified, avoids insurance company an innocent person from suffering a loss, improves the efficiency of claim processing.
The embodiment of the present invention provides a kind of Insurance Fraud recognition methods based on teledata, including：
The personnel concerning the case between any two mutual is extracted from personnel concerning the case's teledata of insurance case to be identified
Contact data；Wherein, the data that connect each other include at least one in talk times, the duration of call or short message length；
According to the personnel concerning the case between any two connect data each other, using the first data model being pre-configured with, meter
Calculate the coefficient correlation of the personnel concerning the case between any two；
According to the coefficient correlation, judge the personnel concerning the case whether close relation；
According to the personnel concerning the case whether close relation, judge the insurance case to be identified whether be fraud case.
Implement the embodiment of the present invention, have the advantages that：
Insurance Fraud recognition methods provided in an embodiment of the present invention based on teledata, obtained according to teledata case-involving
The coefficient correlation of personnel between any two, close relation is judged whether according to coefficient correlation and then judges to insure whether case is fraud
Case.According only to the data analysis of insurance case whether it is fraud case compared to prior art, the embodiment of the present invention make use of
Degree in close relations between teledata analysis personnel concerning the case, if the close relation between personnel concerning the case, there is high confidence level
Personnel concerning the case is identified as clique, and then judges fraud case, sends alarm to insurance company in time, avoids insurance company not guilty
Suffer a loss, and the case for being judged as non-fraud can immediately enter claim adjustment program, improve the efficiency of claim processing,
Reduce resource, human cost that insurance company is used to investigate.
Further, it is described according to the coefficient correlation, judge the personnel concerning the case whether close relation, specifically include：
Calculate the summation of the coefficient correlation；
If the summation of the coefficient correlation is more than or equal to default first threshold, confirm that personnel concerning the case's relation is tight
It is close；
If the summation of the coefficient correlation is less than default first threshold, it is not that relation is tight to confirm the personnel concerning the case
It is close.
Further, it is described according to the personnel concerning the case whether close relation, judge the insurance case to be identified
Before whether being fraud case, in addition to：
Fisrt feature variable is extracted from the initial data of insurance case to be identified and personnel concerning the case's teledata；Its
In, the fisrt feature variable has multiple；
According to the fisrt feature variable, using the second data model being pre-configured with, the insurance to be identified is calculated
The probability of cheating of case；Wherein, second data model is the initial data and personnel concerning the case's telecommunications with multiple history cases
Data obtain as sample training；
Then it is described according to the personnel concerning the case whether close relation, judge whether the insurance case to be identified is fraud
Case, it is specially：
With reference to the personnel concerning the case whether close relation conclusion and the probability of cheating, judge the insurance to be identified
Whether case is fraud case.
Further, it is described according to the personnel concerning the case whether close relation, judge the insurance case to be identified
Before whether being fraud case, in addition to：
If personnel concerning the case's close relation, according to the coefficient correlation, the core person in personnel concerning the case is obtained；And
The blacklist being pre-configured with is inquired about, judges whether the core person belongs to blacklist personnel；
The personnel concerning the case with reference to described in whether close relation conclusion and the probability of cheating, judge described to be identified
Whether insurance case is fraud case, is specifically included：
According to assignment table set in advance, be the personnel concerning the case whether the conclusion assignment of close relation, obtain clique and take advantage of
Cheat factor；
It is the conclusion the assignment whether core person belongs to blacklist personnel according to assignment table set in advance；Obtain
Personnel cheat factor；
According to weight table set in advance, the clique is cheated factor, each personnel concerning the case fraud factor and take advantage of
Swindleness probability is weighted summation；
According to the operation result of the weighted sum, judge whether the insurance case to be identified is fraud case；
Wherein, the assignment table for recording the personnel concerning the case, whether with corresponding clique cheat by the conclusion of close relation
Whether factor, and the core person belong to the conclusion and corresponding personnel fraud factor of blacklist personnel；The weight table
For recording the weight of clique's fraud factor, personnel cheat the weight of factor and the weight of probability of cheating.
Further, second data model is neural network prediction model.
Correspondingly, the present invention also provides a kind of Insurance Fraud identification device based on teledata, including：
First extraction module, for extracting the case-involving people from personnel concerning the case's teledata of insurance case to be identified
Member between any two connect data each other；Wherein, the data that connect each other include talk times, the duration of call or short message length
At least one of in；
First computing module, for according to the personnel concerning the case between any two connect data each other, using being pre-configured with
The first data model, calculate the coefficient correlation of the personnel concerning the case between any two；
First judge module, for according to the coefficient correlation, judge the personnel concerning the case whether close relation；
Second judge module, for according to the personnel concerning the case whether close relation, judge the insurance case to be identified
Whether example is fraud case.
Insurance Fraud identification device provided in an embodiment of the present invention based on teledata, obtained according to teledata case-involving
The coefficient correlation of personnel between any two, close relation is judged whether according to coefficient correlation and then judges to insure whether case is fraud
Case.According only to the data analysis of insurance case whether it is fraud case compared to prior art, the embodiment of the present invention make use of
Degree in close relations between teledata analysis personnel concerning the case, if close relation between personnel concerning the case, there is high confidence level will
Personnel concerning the case is identified as clique, and then judges fraud case, sends alarm to insurance company in time, avoids insurance company is not guilty from meeting with
Suffer a loss, and the case for being judged as non-fraud can immediately enter claim adjustment program, improve the efficiency of claim processing, drop
Low insurance company is used for resource, the human cost investigated.
Further, the first judge module includes：
Sum calculation unit, for calculating the summation of the coefficient correlation；
First confirmation unit, if the summation for the coefficient correlation is more than or equal to default first threshold, confirm
The personnel concerning the case is close relation；
If the summation of the second confirmation unit coefficient correlation is less than default first threshold, the personnel concerning the case is confirmed
It is not close relation.
Further, the Insurance Fraud identification device also includes：
Second extraction module, for being extracted from the initial data of insurance case to be identified and personnel concerning the case's teledata
Fisrt feature variable；Wherein, the fisrt feature variable has multiple；
Second computing module, for according to the fisrt feature variable, using the second data model being pre-configured with, calculating
The probability of cheating of the insurance case to be identified；Wherein, second data model is the original number with multiple history cases
According to what is obtained with personnel concerning the case's teledata as sample training；
Then second judge module be specifically used for reference to the personnel concerning the case whether close relation conclusion and described take advantage of
Probability is cheated, judges whether the insurance case to be identified is fraud case.
Further, the Insurance Fraud identification device also includes：
3rd judge module, for inquiring about the blacklist being pre-configured with, judge whether each personnel concerning the case belongs to blacklist people
Member；
Second judge module specifically includes：
First assignment unit, for according to assignment table set in advance, be the personnel concerning the case whether the knot of close relation
By assignment, clique's fraud factor is obtained；
Second assignment unit, for according to assignment table set in advance, whether belonging to blacklist people for each personnel concerning the case
The conclusion assignment of member；Obtain the fraud factor of each personnel concerning the case；
Sum unit, for according to weight table set in advance, cheating factor to the clique, each personnel concerning the case takes advantage of
Swindleness factor and probability of cheating are weighted summation；
Conclusion unit, for the operation result according to the weighted sum, whether judge the insurance case to be identified
To cheat case；
Wherein, the assignment table for recording the personnel concerning the case, whether with corresponding clique cheat by the conclusion of close relation
Whether factor, and the core person belong to the conclusion and corresponding personnel fraud factor of blacklist personnel；The weight table
For recording the weight of clique's fraud factor, personnel cheat the weight of factor and the weight of probability of cheating.
Further, second data model is neural network prediction model.
Brief description of the drawings
Fig. 1 is the flow chart for the Insurance Fraud recognition methods based on teledata that the embodiment of the present invention one provides；
Fig. 2 is the flow chart for the Insurance Fraud recognition methods based on teledata that the embodiment of the present invention two provides；
Fig. 3 is the close relation detects schematic diagram exported in the embodiment of the present invention two；
Fig. 4 is the structure money figure for the Insurance Fraud identification device based on teledata that the embodiment of the present invention three provides.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
Embodiment one
It is the flow chart for the Insurance Fraud recognition methods based on teledata that the embodiment of the present invention one provides referring to Fig. 1；
This method includes：
S101, extract from personnel concerning the case's teledata of insurance case to be identified the personnel concerning the case between any two
Connect data each other；Wherein, the data that connect each other include at least one in talk times, the duration of call or short message length
；
S102, according to the personnel concerning the case between any two connect data each other, using the first data mould being pre-configured with
Type, calculate the coefficient correlation of the personnel concerning the case between any two；
S103, according to the coefficient correlation, judge the personnel concerning the case whether close relation；
S104, according to the personnel concerning the case whether close relation, judge whether the insurance case to be identified is fraud
Case.
Specifically, insurance case personnel concerning the case has multiple, and by taking vehicle insurance as an example, personnel concerning the case may include driver, insure
People, insurant, beneficiary and car owner.For above personnel concerning the case, if it is in close relations between each other, there is very big probability to be
Close relation.Embodiment one utilize as rule, by extract teledata carry out data processing, solve prior art without
The technical problem whether method accurate judgement insurance case cheats.
The Insurance Fraud recognition methods based on teledata that embodiment one provides, personnel concerning the case is obtained according to teledata
Coefficient correlation between any two, close relation is judged whether according to coefficient correlation and then judges to insure whether case is case of victimization
Example.According only to the data analysis of insurance case whether it is fraud case compared to prior art, the embodiment of the present invention make use of electricity
Degree in close relations between letter data analysis personnel concerning the case, if close relation between personnel concerning the case, has high confidence level to relate to
Case giver identification is clique, so judge fraud case, send alarm to insurance company in time, avoid insurance company an innocent person by
Loss, and the case for being judged as non-fraud can immediately enter claim adjustment program, improve the efficiency of claim processing, reduce
Insurance company is used for resource, the human cost investigated.
Embodiment two
It is the flow chart for the Insurance Fraud recognition methods based on teledata that the embodiment of the present invention two provides referring to Fig. 2；
This method includes：
S201, extract from personnel concerning the case's teledata of insurance case to be identified the personnel concerning the case between any two
Connect data each other；Wherein, the data that connect each other include at least one in talk times, the duration of call or short message length
；
S202, according to the personnel concerning the case between any two connect data each other, using the first data mould being pre-configured with
Type, calculate the coefficient correlation of the personnel concerning the case between any two；
S203, according to the coefficient correlation, judge the personnel concerning the case whether close relation；
S204, fisrt feature change is extracted from the initial data of insurance case to be identified and personnel concerning the case's teledata
Amount；Wherein, the fisrt feature variable has multiple；
S205, according to the fisrt feature variable, using the second data model being pre-configured with, calculate described to be identified
Insure the probability of cheating of case；Wherein, second data model is the initial data with multiple history cases and personnel concerning the case
Teledata obtains as sample training；
If S206, personnel concerning the case's close relation, according to the coefficient correlation, the core people in personnel concerning the case is obtained
Member；And the blacklist being pre-configured with is inquired about, judge whether the core person belongs to blacklist personnel；
S207, judge whether the insurance case to be identified is fraud case.
As preferred embodiment, whether first data model can be with known to multipair between related personnel
Relation and its connect the machine learning model that data train each other, can be according to personnel concerning the case two-by-two using the first data model
Between the data that connect each other calculate its related probability, i.e. coefficient correlation.
For the present embodiment compared with embodiment one, the content auxiliary for adding two dimensions carries out Insurance Fraud identification.I.e. from
Degree in close relations between personnel concerning the case, the probability of cheating based on data model calculate, personnel concerning the case whether blacklist personnel
Whether three dimensional analysis insurance case to be identified is fraud case.
For first dimension, the degree in close relations between personnel concerning the case, the present embodiment is led to by telecommunication user history
The communication contact information of the means such as words, short message, flow, excavates the relevance between multi-party personnel concerning the case.Specifically, according to telecommunications
The interactive information such as the history call of user, short message, online, interaction feature between user is extracted (as conversed and transmitting-receiving message frequency, logical
Talk about duration, short message length, traffic frequencies etc.), so as to spend tax closely to the interaction user using the data model being pre-configured with
Value, obtains coefficient correlation.After the procedure, it can show complete personnel concerning the case's with the close detects schematic diagram of output relation
Relational network supplies the insurer officer of responsible fraud identification to check, close relation detects schematic diagram as shown in Figure 3.In the figure
In, the connecting line of the interactive relation of personnel concerning the case between node, it is every connection according to coefficient correlation using personnel concerning the case as node
Line assigns weight, to characterize interaction degree distribution closely between user.
After coefficient correlation is obtained, step S203, according to the coefficient correlation, judging the personnel concerning the case, whether relation is tight
It is close, specifically include：
Calculate the summation of the coefficient correlation；
If the summation of the coefficient correlation is more than or equal to default first threshold, it is relation to confirm the personnel concerning the case
Closely；
If the summation of the coefficient correlation is less than default first threshold, it is not that relation is tight to confirm the personnel concerning the case
It is close.
In above-mentioned steps S203 concrete scheme, the summation of the coefficient correlation is calculated, that is, the relation to Fig. 3
Each connecting line of close detects schematic diagram, is multiplied by its weight, then is superimposed, and obtains the journey in close relations of whole group personnel concerning the case
Degree.By setting rational threshold value, can draw personnel concerning the case whether the conclusion of close relation.
For second dimension, the probability of cheating based on data model calculates, by the way that telecommunication user history is conversed, paid
The data such as expense, arrearage and the claim case related data of insurance company's offer, partly have confirmed that the case data of fraud is combined,
Build the second data model of depth fraud detection, various changeable fraud pattern when depth excavates customer claim.Wherein, telecommunications
Data include at least one following：It is history log (calling number, communication beginning and ending time, data traffic use), short
Letter information (sender's number, addressee's number, short message length, transmit and collection of letters place etc.), internet information (accounts information, industry
Business tine length that business stream beginning and ending time, user receive, business objective, categories of websites etc.), bill record (account balance, pays
Take record, ARPU etc.), personal information (age, sex, occupation, home address etc.), using set meal situation, social situation, subscription
Service log, complain record.Insuring the initial data of case includes at least one following：Report a case to the security authorities the time, place of reporting a case to the security authorities, case
It is related to personnel's telephone number, case is related to information of vehicles, if fraud case label.
As a preferred embodiment, using the initial data of multiple history cases and personnel concerning the case's teledata as
Sample, by characteristic vector pickup algorithm, representative fisrt feature variable is extracted, and be based in this, as input structure
The forecast model of neural network algorithm, as the second data model.Neural network model is nonlinear by multiple essential characteristics
Combination, has powerful self-learning capability, be adapted to processing needs and meanwhile consider many factors and condition, it is inaccurate and fuzzy
Information-processing problem, it can fully approach the non-linear relation of complexity.If there is error in output valve, can be anti-by error with desired value
Network is trained to propagation algorithm, such as adjusts weighting parameter at each node, until model is restrained or reaches certain accurate
Degree.Meanwhile neural network model can train according to newest data so that model has adaptivity, can be preferably
Validity feature is caught, improves fraud detection success rate.The output result of the dimension, one is embodied between (0,1)
Floating number, represent the probability of cheating of insurance case to be identified.
For the 3rd dimension, if i.e. step S206, personnel concerning the case's close relation, according to the coefficient correlation,
Obtain the core person in personnel concerning the case；And the blacklist being pre-configured with is inquired about, judge whether the core person belongs to black name
Single member, it make use of and establish the high risk of fraud user of blacklist screening, quick inquiry in real time is realized, if core member is black name
Single member then illustrates that the clique is potential fraud clique, and designing corresponding judgement scheme raising in step S207 is identified as case of victimization
The possibility of example.Wherein, the core person obtained in personnel concerning the case can be that the coefficient correlation being related to each personnel concerning the case is entered
Row summation, using the maximum personnel concerning the case of the coefficient correlation sum being related to as core person.
As a preferred embodiment, the present embodiment utilizes the existing user's history call note of telecommunication service system
The interaction data information such as record, short message, online, user's portrait of telecommunication user is built, so as to filter out high risk of fraud user, is built
Stand and safeguard fraud blacklist, to inquiry during fraud detection.Telecommunication service data are primarily based on, system will extract wherein
Important user profile, as user's portrait dimension.It includes：A. identity speciality；B. social quality；C. contractual capacity；D. go
For preference.User's portrait of telecommunication user is built using these dimensions as mode input.Drawn a portrait for user, system is by each dimension
Number of degrees value, on the basis of each dimension scoring is obtained using telecommunication user historical data, build the 3rd data model.
Preferably, the 3rd data model is random forest machine learning model.Random forest is numerous machine learning classifications
In device, a very outstanding grader of modeling effect.It considers more decision trees by combining, and forms decision tree " forest ",
Such Fusion Model substantially increases the accuracy rate of prediction.Random forest utilizes the record x being made up of in training set indexm=
[x1,...,xi,...,xj,...,xN], and the tag along sort y definedm={ non-black list user=0, black list user
=1 }, a series of decision tree f is generated using the stochastic subspace of feature spacet,1≤t≤T.Model predication value is random gloomy
The average value of all decision tree predicted values in woods, i.e.,：
Y herein, as telecommunication user are the probable value of black list user.By setting rational threshold value, can therefrom sieve
The user of high risk of fraud is selected, to inquiry during fraud detection.
That is the collocation method of blacklist includes：
According to the second feature variable of each user obtained from telecommunication user database, using the 3rd data model, meter
Calculate the probable value that each user is blacklist personnel；Wherein, the second feature variable is multiple；
The user that blacklist personnel probable value is more than or equal to Second Threshold is chosen, in blacklisting it.
In another embodiment, can also match somebody with somebody directly using personnel concerning the case's list in known Insurance Fraud case
Put blacklist.
After the Data Management Analysis result of three dimensions is obtained, the insurance case to be identified step S207, is judged
Whether it is fraud case, specifically includes：
According to assignment table set in advance, be the personnel concerning the case whether the conclusion assignment of close relation, obtain clique and take advantage of
Cheat factor；
It is the conclusion the assignment whether core person belongs to blacklist personnel according to assignment table set in advance；Obtain
Personnel cheat factor；
According to weight table set in advance, factor is cheated the clique, personnel cheat factor and probability of cheating is carried out
Weighted sum；
According to the operation result of the weighted sum, judge whether the insurance case to be identified is fraud case.
Wherein, the assignment table for recording the personnel concerning the case, whether with corresponding clique cheat by the conclusion of close relation
Whether factor, and the core person belong to the conclusion and corresponding personnel fraud factor of blacklist personnel；The weight table
For recording the weight of clique's fraud factor, personnel cheat the weight of factor and the weight of probability of cheating.
It should be noted that the present embodiment judges the guarantor to be identified using the Data Analysis Services result of three dimensions
Whether dangerous case is fraud case, an only preferred embodiment, in other embodiments, can be entered only with first dimension
Row analysis, may also be used in which 2 dimensions or is combined above-mentioned 3 dimensions with other dimensions and analyzed, do not departing from this hair
On the premise of bright principle, these embodiments are all considered as protection scope of the present invention.For example, for only relating to first dimension and
The analysis of two dimensions, it is described according to the personnel concerning the case whether close relation, whether judge the insurance case to be identified
To cheat case, it is specially：With reference to the personnel concerning the case whether close relation conclusion and the probability of cheating, treat described in judgement
Whether the insurance case of identification is fraud case.
The Insurance Fraud recognition methods based on teledata that embodiment two provides, personnel concerning the case is obtained according to teledata
Coefficient correlation between any two, close relation is judged whether according to coefficient correlation and then judges to insure whether case is case of victimization
Example.According only to the data analysis of insurance case whether it is fraud case compared to prior art, the embodiment of the present invention make use of electricity
Degree in close relations between letter data analysis personnel concerning the case, if close relation between personnel concerning the case, has high confidence level to relate to
Case giver identification is clique, so judge fraud case, send alarm to insurance company in time, avoid insurance company an innocent person by
Loss, and the case for being judged as non-fraud can immediately enter claim adjustment program, improve the efficiency of claim processing, reduce
Insurance company is used for resource, the human cost investigated.
Embodiment three
It is the structure money for the Insurance Fraud identification device based on teledata that the embodiment of the present invention three provides referring to Fig. 4
Figure.
Correspondingly, the present invention also provides a kind of Insurance Fraud identification device based on teledata, including：
First extraction module 401, for being related to described in the extraction from personnel concerning the case's teledata of insurance case to be identified
Case personnel between any two connect data each other；Wherein, the data that connect each other include talk times, the duration of call or short message
At least one of in length；
First computing module 402, for according to the personnel concerning the case between any two connect data each other, using matching somebody with somebody in advance
The first data model put, calculate the coefficient correlation of the personnel concerning the case between any two；
First judge module 403, for according to the coefficient correlation, judge the personnel concerning the case whether close relation；
Second judge module 404, for according to the personnel concerning the case whether close relation, judge the insurance to be identified
Whether case is fraud case.
Further, the first judge module includes：
Sum calculation unit, for calculating the summation of the coefficient correlation；
First confirmation unit, if the summation for the coefficient correlation is more than or equal to default first threshold, confirm
Personnel concerning the case's close relation；
If the summation of the second confirmation unit coefficient correlation is less than default first threshold, the personnel concerning the case is confirmed
It is not close relation.
Further, the Insurance Fraud identification device also includes：
Second extraction module, for being extracted from the initial data of insurance case to be identified and personnel concerning the case's teledata
Fisrt feature variable；Wherein, the fisrt feature variable has multiple；
Second computing module, for according to the fisrt feature variable, using the second data model being pre-configured with, calculating
The probability of cheating of the insurance case to be identified；Wherein, second data model is the original number with multiple history cases
According to what is obtained with personnel concerning the case's teledata as sample training；
Then second judge module be specifically used for reference to the personnel concerning the case whether close relation conclusion and described take advantage of
Probability is cheated, judges whether the insurance case to be identified is fraud case.
Further, the Insurance Fraud identification device also includes：
3rd judge module, for inquiring about the blacklist being pre-configured with, judge whether each personnel concerning the case belongs to blacklist people
Member；
Second judge module specifically includes：
First assignment unit, for according to assignment table set in advance, be the personnel concerning the case whether the knot of close relation
By assignment, clique's fraud factor is obtained；
Second assignment unit, for according to assignment table set in advance, whether belonging to blacklist people for each personnel concerning the case
The conclusion assignment of member；Obtain the fraud factor of each personnel concerning the case；
Sum unit, for according to weight table set in advance, cheating factor to the clique, each personnel concerning the case takes advantage of
Swindleness factor and probability of cheating are weighted summation；
Conclusion unit, for the operation result according to the weighted sum, whether judge the insurance case to be identified
To cheat case.
Further, second data model is neural network prediction model；3rd data model is random gloomy
Woods data model.
The Insurance Fraud identification device based on teledata that example IV provides, personnel concerning the case is obtained according to teledata
Coefficient correlation between any two, close relation is judged whether according to coefficient correlation and then judges to insure whether case is case of victimization
Example.According only to the data analysis of insurance case whether it is fraud case compared to prior art, the embodiment of the present invention make use of electricity
Degree in close relations between letter data analysis personnel concerning the case, if close relation between personnel concerning the case, has high confidence level to relate to
Case giver identification is clique, so judge fraud case, send alarm to insurance company in time, avoid insurance company an innocent person by
Loss, and the case for being judged as non-fraud can immediately enter claim adjustment program, improve the efficiency of claim processing, reduce
Insurance company is used for resource, the human cost investigated.
One of ordinary skill in the art will appreciate that realize all or part of flow in above-described embodiment method, being can be with
The hardware of correlation is instructed to complete by computer program, described program can be stored in computer read/write memory medium
In, the program is upon execution, it may include such as the flow of the embodiment of above-mentioned each method.Wherein, described storage medium can be magnetic
Dish, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access
Memory, RAM) etc..
Described above is the preferred embodiment of the present invention, it is noted that for those skilled in the art
For, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications are also considered as
Protection scope of the present invention.
Claims (10)
- A kind of 1. Insurance Fraud recognition methods based on teledata, it is characterised in that including：The personnel concerning the case connecting each other between any two is extracted from personnel concerning the case's teledata of insurance case to be identified Data；Wherein, the data that connect each other include at least one in talk times, the duration of call or short message length；According to the personnel concerning the case between any two connect data each other, using the first data model being pre-configured with, calculate institute State the coefficient correlation of personnel concerning the case between any two；According to the coefficient correlation, judge the personnel concerning the case whether close relation；According to the personnel concerning the case whether close relation, judge the insurance case to be identified whether be fraud case.
- 2. the Insurance Fraud recognition methods based on teledata as claimed in claim 1, it is characterised in that described in the basis Coefficient correlation, judge the personnel concerning the case whether close relation, specifically include：Calculate the summation of the coefficient correlation；If the summation of the coefficient correlation is more than or equal to default first threshold, personnel concerning the case's close relation is confirmed；If the summation of the coefficient correlation is less than default first threshold, it is not close relation to confirm the personnel concerning the case.
- 3. the Insurance Fraud recognition methods based on teledata as claimed in claim 1 or 2, it is characterised in that at described According to the personnel concerning the case whether close relation, judge the insurance case to be identified whether be fraud case before, in addition to：Fisrt feature variable is extracted from the initial data of insurance case to be identified and personnel concerning the case's teledata；Wherein, institute State fisrt feature variable have it is multiple；According to the fisrt feature variable, using the second data model being pre-configured with, the insurance case to be identified is calculated Probability of cheating；Wherein, second data model is the initial data and personnel concerning the case's teledata with multiple history cases Obtained as sample training；Then it is described according to the personnel concerning the case whether close relation, judge whether the insurance case to be identified is case of victimization Example, it is specially：With reference to the personnel concerning the case whether close relation conclusion and the probability of cheating, judge the insurance case to be identified Whether it is fraud case.
- 4. the Insurance Fraud recognition methods based on teledata as claimed in claim 3, it is characterised in that described according to institute State personnel concerning the case whether close relation, judge the insurance case to be identified whether be fraud case before, in addition to：If personnel concerning the case's close relation, according to the coefficient correlation, the core person in personnel concerning the case is obtained；And inquire about The blacklist being pre-configured with, judges whether the core person belongs to blacklist personnel；The personnel concerning the case with reference to described in whether close relation conclusion and the probability of cheating, judge the insurance to be identified Whether case is fraud case, is specifically included：According to assignment table set in advance, be the personnel concerning the case whether the conclusion assignment of close relation, obtain clique's fraud because Number；It is the conclusion the assignment whether core person belongs to blacklist personnel according to assignment table set in advance；Obtain personnel Cheat factor；According to weight table set in advance, factor is cheated the clique, personnel cheat factor and probability of cheating is weighted Summation；According to the operation result of the weighted sum, judge whether the insurance case to be identified is fraud case；Wherein, the assignment table be used to recording the personnel concerning the case whether the conclusion of close relation and corresponding clique cheat because Whether number, and the core person belong to the conclusion and corresponding personnel fraud factor of blacklist personnel；The weight table is used In the weight, the weight of personnel's fraud factor and the weight of probability of cheating of recording clique's fraud factor.
- 5. the Insurance Fraud recognition methods based on teledata as claimed in claim 4, it is characterised in that second data Model is neural network prediction model.
- A kind of 6. Insurance Fraud identification device based on teledata, it is characterised in that including：First extraction module, for extracting the personnel concerning the case two from personnel concerning the case's teledata of insurance case to be identified Data are connected each other between two；Wherein, the data that connect each other are included in talk times, the duration of call or short message length At least one of；First computing module, for according to the personnel concerning the case between any two connect data each other, using be pre-configured with One data model, calculate the coefficient correlation of the personnel concerning the case between any two；First judge module, for according to the coefficient correlation, judge the personnel concerning the case whether close relation；Second judge module, for according to the personnel concerning the case whether close relation, judge that the insurance case to be identified is No is fraud case.
- 7. the Insurance Fraud identification device based on teledata as claimed in claim 6, it is characterised in that the first judge module Including：Sum calculation unit, for calculating the summation of the coefficient correlation；First confirmation unit, if the summation for the coefficient correlation is more than or equal to default first threshold, confirm described in Personnel concerning the case's close relation；If the summation of the second confirmation unit coefficient correlation is less than default first threshold, confirm that the personnel concerning the case is not Close relation.
- 8. the Insurance Fraud identification device based on teledata as claimed in claims 6 or 7, it is characterised in that the insurance Fraud identification device also includes：Second extraction module, for extracting first from the initial data of insurance case to be identified and personnel concerning the case's teledata Characteristic variable；Wherein, the fisrt feature variable has multiple；Second computing module, for according to the fisrt feature variable, using the second data model being pre-configured with, described in calculating The probability of cheating of insurance case to be identified；Wherein, second data model be with the initial data of multiple history cases and Personnel concerning the case's teledata obtains as sample training；Then second judge module be specifically used for reference to the personnel concerning the case whether close relation conclusion and the fraud it is general Rate, judge whether the insurance case to be identified is fraud case.
- 9. the Insurance Fraud identification device based on teledata as claimed in claim 8, it is characterised in that the Insurance Fraud Identification device also includes：3rd judge module, if for personnel concerning the case's close relation, according to the coefficient correlation, obtain in personnel concerning the case Core person；And the blacklist being pre-configured with is inquired about, judge whether the core person belongs to blacklist personnel；Second judge module specifically includes：First assignment unit, for being that the personnel concerning the case whether assign by the conclusion of close relation according to assignment table set in advance Value, obtain clique's fraud factor；Second assignment unit, for according to assignment table set in advance, whether belonging to blacklist personnel's for the core person Conclusion assignment, obtain personnel and cheat factor；Sum unit, for according to weight table set in advance, cheating the clique factor, personnel cheat factor and fraud Probability is weighted summation；Conclusion unit, for the operation result according to the weighted sum, judge whether the insurance case to be identified is to take advantage of Fraud case example；Wherein, the assignment table be used to recording the personnel concerning the case whether the conclusion of close relation and corresponding clique cheat because Whether number, and the core person belong to the conclusion and corresponding personnel fraud factor of blacklist personnel；The weight table is used In the weight, the weight of personnel's fraud factor and the weight of probability of cheating of recording clique's fraud factor.
- 10. the Insurance Fraud identification device based on teledata as claimed in claim 9, it is characterised in that second number It is neural network prediction model according to model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710694259.4A CN107403326A (en) | 2017-08-14 | 2017-08-14 | A kind of Insurance Fraud recognition methods and device based on teledata |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710694259.4A CN107403326A (en) | 2017-08-14 | 2017-08-14 | A kind of Insurance Fraud recognition methods and device based on teledata |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107403326A true CN107403326A (en) | 2017-11-28 |
Family
ID=60398232
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710694259.4A Pending CN107403326A (en) | 2017-08-14 | 2017-08-14 | A kind of Insurance Fraud recognition methods and device based on teledata |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107403326A (en) |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108197795A (en) * | 2017-12-28 | 2018-06-22 | 杭州优行科技有限公司 | The account recognition methods of malice group, device, terminal and storage medium |
| CN108257033A (en) * | 2018-01-12 | 2018-07-06 | 中国平安人寿保险股份有限公司 | A kind of declaration form analysis method, device, terminal device and storage medium |
| CN108334647A (en) * | 2018-04-12 | 2018-07-27 | 阿里巴巴集团控股有限公司 | Data processing method, device, equipment and the server of Insurance Fraud identification |
| CN108364233A (en) * | 2018-01-12 | 2018-08-03 | 中国平安人寿保险股份有限公司 | A kind of declaration form methods of risk assessment, device, terminal device and storage medium |
| CN108460689A (en) * | 2018-01-12 | 2018-08-28 | 中国平安人寿保险股份有限公司 | A kind of declaration form analysis method, device, terminal device and storage medium |
| CN108470312A (en) * | 2018-02-07 | 2018-08-31 | 中国平安人寿保险股份有限公司 | Analysis method, device, storage medium and the terminal for case of settling a claim |
| CN108734479A (en) * | 2018-04-12 | 2018-11-02 | 阿里巴巴集团控股有限公司 | Data processing method, device, equipment and the server of Insurance Fraud identification |
| CN109118376A (en) * | 2018-08-14 | 2019-01-01 | 平安医疗健康管理股份有限公司 | Medical insurance premium calculation principle method, apparatus, computer equipment and storage medium |
| CN109214944A (en) * | 2018-08-28 | 2019-01-15 | 北京费马科技有限公司 | Junk user recognition methods and application based on social graph |
| CN109544379A (en) * | 2018-10-30 | 2019-03-29 | 平安科技（深圳）有限公司 | Flat type determines method and Related product |
| CN109614496A (en) * | 2018-09-27 | 2019-04-12 | 长威信息科技发展股份有限公司 | A kind of minimum living discrimination method of knowledge based map |
| CN109658222A (en) * | 2018-10-16 | 2019-04-19 | 深圳壹账通智能科技有限公司 | Risk analysis method, device, equipment and computer readable storage medium |
| CN109816519A (en) * | 2019-01-25 | 2019-05-28 | 宜人恒业科技发展（北京）有限公司 | A kind of recognition methods of fraud clique, device and equipment |
| CN110008346A (en) * | 2019-01-24 | 2019-07-12 | 平安科技（深圳）有限公司 | Blacklist conducts map construction method, apparatus, computer equipment and storage medium |
| CN110008220A (en) * | 2019-01-24 | 2019-07-12 | 平安科技（深圳）有限公司 | Blacklist conducts map attenuation coefficient acquisition methods, device and computer equipment |
| CN110008347A (en) * | 2019-01-24 | 2019-07-12 | 平安科技（深圳）有限公司 | Blacklist conducts extending method, device, computer equipment and storage medium |
| CN110175928A (en) * | 2019-05-14 | 2019-08-27 | 安泰保险科技(深圳)有限公司 | A kind of insurance that the intentional manufacture vehicle of identification mutually touches insurance risk is counter to cheat method |
| CN110175927A (en) * | 2019-05-14 | 2019-08-27 | 安泰保险科技(深圳)有限公司 | It is a kind of to be in danger that the insurance stealthily substituted is counter to cheat method based on carrier data identification drunk driving |
| CN110223086A (en) * | 2019-05-30 | 2019-09-10 | 珠海市岭南大数据研究院 | Vehicle insurance is reported a case to the security authorities recognition methods, device and the computer equipment of information |
| CN110827159A (en) * | 2019-11-11 | 2020-02-21 | 上海交通大学 | Financial medical insurance fraud early warning method, device and terminal based on relational graph |
| CN111143626A (en) * | 2019-12-25 | 2020-05-12 | 深圳云天励飞技术有限公司 | Group partner identification method, device, equipment and computer readable storage medium |
| CN111428217A (en) * | 2020-04-12 | 2020-07-17 | 中信银行股份有限公司 | Method and device for identifying cheat group, electronic equipment and computer readable storage medium |
| CN111861762A (en) * | 2020-07-28 | 2020-10-30 | 贵州力创科技发展有限公司 | Data processing method and system for anti-fraud recognition of vehicle insurance |
| CN114170030A (en) * | 2021-12-08 | 2022-03-11 | 北京百度网讯科技有限公司 | Method, device, electronic equipment and medium for remote damage assessment of vehicle |
| CN115034918A (en) * | 2022-08-09 | 2022-09-09 | 太平金融科技服务(上海)有限公司深圳分公司 | Ganged case identification method, ganged case identification device, ganged case identification computer equipment, storage medium and product |
| CN111861762B (en) * | 2020-07-28 | 2024-04-26 | 贵州力创科技发展有限公司 | Data processing method and system for identifying anti-fraud safety of vehicle |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102393843A (en) * | 2011-06-29 | 2012-03-28 | 广州市动景计算机科技有限公司 | Method and system for establishing relational network of user by using communication information of mobile terminal |
| CN103763152A (en) * | 2014-01-07 | 2014-04-30 | 中国移动（深圳）有限公司 | Method and system for multi-dimensionally monitoring telecommunication fraudulent conduct |
| CN103810637A (en) * | 2013-12-17 | 2014-05-21 | 深圳市般若计算机系统有限公司 | Motor vehicle insurance fraud detecting method and system |
| CN106600423A (en) * | 2016-11-18 | 2017-04-26 | 云数信息科技（深圳）有限公司 | Machine learning-based car insurance data processing method and device and car insurance fraud identification method and device |
-
2017
- 2017-08-14 CN CN201710694259.4A patent/CN107403326A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102393843A (en) * | 2011-06-29 | 2012-03-28 | 广州市动景计算机科技有限公司 | Method and system for establishing relational network of user by using communication information of mobile terminal |
| CN103810637A (en) * | 2013-12-17 | 2014-05-21 | 深圳市般若计算机系统有限公司 | Motor vehicle insurance fraud detecting method and system |
| CN103763152A (en) * | 2014-01-07 | 2014-04-30 | 中国移动（深圳）有限公司 | Method and system for multi-dimensionally monitoring telecommunication fraudulent conduct |
| CN106600423A (en) * | 2016-11-18 | 2017-04-26 | 云数信息科技（深圳）有限公司 | Machine learning-based car insurance data processing method and device and car insurance fraud identification method and device |
Non-Patent Citations (1)
| Title |
|---|
| 清华大学数据科学研究院: ""RONG"系列论坛之四：保险理赔反欺诈之大数据应用(林晟)", 《HTTP://WWW.IDS.TSINGHUA.EDU.CN/INDEX.PHP?P=3277》 * |
Cited By (39)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108197795A (en) * | 2017-12-28 | 2018-06-22 | 杭州优行科技有限公司 | The account recognition methods of malice group, device, terminal and storage medium |
| CN108460689B (en) * | 2018-01-12 | 2021-07-16 | 中国平安人寿保险股份有限公司 | Policy analysis method and device, terminal equipment and storage medium |
| CN108257033A (en) * | 2018-01-12 | 2018-07-06 | 中国平安人寿保险股份有限公司 | A kind of declaration form analysis method, device, terminal device and storage medium |
| CN108364233A (en) * | 2018-01-12 | 2018-08-03 | 中国平安人寿保险股份有限公司 | A kind of declaration form methods of risk assessment, device, terminal device and storage medium |
| CN108460689A (en) * | 2018-01-12 | 2018-08-28 | 中国平安人寿保险股份有限公司 | A kind of declaration form analysis method, device, terminal device and storage medium |
| CN108470312A (en) * | 2018-02-07 | 2018-08-31 | 中国平安人寿保险股份有限公司 | Analysis method, device, storage medium and the terminal for case of settling a claim |
| CN108470312B (en) * | 2018-02-07 | 2021-07-06 | 中国平安人寿保险股份有限公司 | Method and device for analyzing claim case, storage medium and terminal |
| CN108334647A (en) * | 2018-04-12 | 2018-07-27 | 阿里巴巴集团控股有限公司 | Data processing method, device, equipment and the server of Insurance Fraud identification |
| CN108734479A (en) * | 2018-04-12 | 2018-11-02 | 阿里巴巴集团控股有限公司 | Data processing method, device, equipment and the server of Insurance Fraud identification |
| WO2019196552A1 (en) * | 2018-04-12 | 2019-10-17 | 阿里巴巴集团控股有限公司 | Data processing method, apparatus and device for insurance fraud identification, and server |
| TWI686760B (en) * | 2018-04-12 | 2020-03-01 | 香港商阿里巴巴集團服務有限公司 | Data processing method, device, equipment and server for insurance fraud identification |
| CN109118376A (en) * | 2018-08-14 | 2019-01-01 | 平安医疗健康管理股份有限公司 | Medical insurance premium calculation principle method, apparatus, computer equipment and storage medium |
| CN109214944A (en) * | 2018-08-28 | 2019-01-15 | 北京费马科技有限公司 | Junk user recognition methods and application based on social graph |
| CN109214944B (en) * | 2018-08-28 | 2022-03-11 | 北京蚂蚁云金融信息服务有限公司 | Social graph-based junk user identification method and device |
| CN109614496B (en) * | 2018-09-27 | 2022-06-17 | 长威信息科技发展股份有限公司 | Low security identification method based on knowledge graph |
| CN109614496A (en) * | 2018-09-27 | 2019-04-12 | 长威信息科技发展股份有限公司 | A kind of minimum living discrimination method of knowledge based map |
| CN109658222A (en) * | 2018-10-16 | 2019-04-19 | 深圳壹账通智能科技有限公司 | Risk analysis method, device, equipment and computer readable storage medium |
| CN109544379A (en) * | 2018-10-30 | 2019-03-29 | 平安科技（深圳）有限公司 | Flat type determines method and Related product |
| CN110008346B (en) * | 2019-01-24 | 2024-05-03 | 平安科技（深圳）有限公司 | Blacklist conduction map construction method, device, computer equipment and storage medium |
| CN110008220B (en) * | 2019-01-24 | 2024-03-15 | 平安科技（深圳）有限公司 | Method and device for obtaining attenuation coefficient of blacklist conduction spectrum and computer equipment |
| CN110008220A (en) * | 2019-01-24 | 2019-07-12 | 平安科技（深圳）有限公司 | Blacklist conducts map attenuation coefficient acquisition methods, device and computer equipment |
| CN110008346A (en) * | 2019-01-24 | 2019-07-12 | 平安科技（深圳）有限公司 | Blacklist conducts map construction method, apparatus, computer equipment and storage medium |
| CN110008347B (en) * | 2019-01-24 | 2024-05-03 | 平安科技（深圳）有限公司 | Blacklist conduction expansion method, device, computer equipment and storage medium |
| CN110008347A (en) * | 2019-01-24 | 2019-07-12 | 平安科技（深圳）有限公司 | Blacklist conducts extending method, device, computer equipment and storage medium |
| CN109816519A (en) * | 2019-01-25 | 2019-05-28 | 宜人恒业科技发展（北京）有限公司 | A kind of recognition methods of fraud clique, device and equipment |
| CN110175928A (en) * | 2019-05-14 | 2019-08-27 | 安泰保险科技(深圳)有限公司 | A kind of insurance that the intentional manufacture vehicle of identification mutually touches insurance risk is counter to cheat method |
| CN110175927A (en) * | 2019-05-14 | 2019-08-27 | 安泰保险科技(深圳)有限公司 | It is a kind of to be in danger that the insurance stealthily substituted is counter to cheat method based on carrier data identification drunk driving |
| CN110223086A (en) * | 2019-05-30 | 2019-09-10 | 珠海市岭南大数据研究院 | Vehicle insurance is reported a case to the security authorities recognition methods, device and the computer equipment of information |
| CN110827159B (en) * | 2019-11-11 | 2023-11-03 | 上海交通大学 | Financial medical insurance fraud early warning method, device and terminal based on relation diagram |
| CN110827159A (en) * | 2019-11-11 | 2020-02-21 | 上海交通大学 | Financial medical insurance fraud early warning method, device and terminal based on relational graph |
| CN111143626A (en) * | 2019-12-25 | 2020-05-12 | 深圳云天励飞技术有限公司 | Group partner identification method, device, equipment and computer readable storage medium |
| CN111428217B (en) * | 2020-04-12 | 2023-07-28 | 中信银行股份有限公司 | Fraudulent party identification method, apparatus, electronic device and computer readable storage medium |
| CN111428217A (en) * | 2020-04-12 | 2020-07-17 | 中信银行股份有限公司 | Method and device for identifying cheat group, electronic equipment and computer readable storage medium |
| CN111861762A (en) * | 2020-07-28 | 2020-10-30 | 贵州力创科技发展有限公司 | Data processing method and system for anti-fraud recognition of vehicle insurance |
| CN111861762B (en) * | 2020-07-28 | 2024-04-26 | 贵州力创科技发展有限公司 | Data processing method and system for identifying anti-fraud safety of vehicle |
| CN114170030A (en) * | 2021-12-08 | 2022-03-11 | 北京百度网讯科技有限公司 | Method, device, electronic equipment and medium for remote damage assessment of vehicle |
| CN114170030B (en) * | 2021-12-08 | 2023-09-26 | 北京百度网讯科技有限公司 | Method, apparatus, electronic device and medium for remote damage assessment of vehicle |
| CN115034918B (en) * | 2022-08-09 | 2023-01-10 | 太平金融科技服务(上海)有限公司深圳分公司 | Ganged case identification method, ganged case identification device, ganged case identification computer equipment, storage medium and product |
| CN115034918A (en) * | 2022-08-09 | 2022-09-09 | 太平金融科技服务(上海)有限公司深圳分公司 | Ganged case identification method, ganged case identification device, ganged case identification computer equipment, storage medium and product |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107403326A (en) | A kind of Insurance Fraud recognition methods and device based on teledata | |
| CN116342259A (en) | Automatic user credit rating method and device, electronic equipment and medium | |
| CN106651570A (en) | System and method for real-time loan approval | |
| CN108230151A (en) | A kind of suspicious transaction detection method, apparatus, equipment and storage medium | |
| CN1296694A (en) | Automated fraud management in transaction-based networks | |
| CN110837963A (en) | Risk control platform construction method based on data, model and strategy | |
| KR101645992B1 (en) | Apparatus for monitering customer complaint and computer-readable medium thereof | |
| CN107835247B (en) | Credit authentication and guarantee system and method | |
| CN110807699B (en) | Overdue event payment collection method and device and computer readable storage medium | |
| CN112767136A (en) | Credit anti-fraud identification method, credit anti-fraud identification device, credit anti-fraud identification equipment and credit anti-fraud identification medium based on big data | |
| CN109254979A (en) | A kind of personal credit evaluation method and system | |
| CN106599200A (en) | Intelligent remote financial data anti-cheating system and anti-cheating method thereof | |
| Zhang et al. | Behavior-based telecommunication churn prediction with neural network approach | |
| CN113724067A (en) | Receiving method, storage medium and device | |
| CN115760332A (en) | Risk prediction method, system, medium and device based on enterprise data analysis | |
| CN106570753A (en) | Method and server for controlling default risk | |
| CN110728570A (en) | Anti-fraud fund analysis method | |
| CN110457601A (en) | The recognition methods and device of social account, storage medium and electronic device | |
| CN111105064A (en) | Method and device for determining suspected information of fraud event | |
| CN113052422A (en) | Wind control model training method and user credit evaluation method | |
| CN111062422A (en) | Method and device for systematic identification of road loan | |
| CN115048472A (en) | Method, device and equipment for intelligently identifying family circle in communication industry | |
| CN112598225A (en) | Evaluation index determination method and apparatus, storage medium, and electronic apparatus | |
| CN113645356A (en) | Fraud telephone identification method and system based on in-network card opening behavior analysis | |
| CN110956471A (en) | Method for analyzing credit investigation data of decoration industry |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171128 |