CN109785174A - The method and apparatus for identifying high risk of fraud - Google Patents
The method and apparatus for identifying high risk of fraud Download PDFInfo
- Publication number
- CN109785174A CN109785174A CN201910105306.6A CN201910105306A CN109785174A CN 109785174 A CN109785174 A CN 109785174A CN 201910105306 A CN201910105306 A CN 201910105306A CN 109785174 A CN109785174 A CN 109785174A
- Authority
- CN
- China
- Prior art keywords
- risk
- area
- high risk
- stay
- length
- 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
Landscapes
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The present invention provides a kind of method, apparatus for identifying high risk of fraud, computer system and computer readable medium.A kind of method identifying high risk of fraud of the invention includes obtaining Claims Resolution case data；Regional case number of packages, geographic classification condition case number of packages, geographic classification conditional average length of stay and geographic classification condition length of stay standard deviation are calculated according to Claims Resolution case data；Determine high-risk area；Determine the high risk class condition for corresponding to high-risk area；Set high risk judgment threshold；According to high-risk area and high risk class condition, high risk case is determined from customer insured.
Description
Technical field
The present invention relates to medical insurance information processing methods, and in particular to a kind of method for identifying high risk of fraud, dress
It sets, computer system and computer readable medium.
Background technique
Under the combination of current internet and insurance, the maximum challenge that internet insurance company encounters is to carry out dinuclear wind
Control reduces high loss ratio, enhances profitability.Specifically, external data quality defect is faced, identification high risk can not be improved
The discrimination rate of client, and largely introduction external data can compress the profit space of short-term accident insurance and health insurance.In this situation
Under, how internet insurance company takes phase using the past client Claims Resolution data accumulated, the client of efficient identification high risk
It answers air control measure to reduce Claims Resolution risk, controls running cost, promote company's profit, be the important process of internet insurance air control.
Summary of the invention
The purpose of the disclosure is data of settling a claim according to the past client that internet insurance company accumulates, and utilizes statistical analysis
Method establishes model, and the medicine of the development degree in area, the rank of hospital and disease is combined rationally to be hospitalized range, sets early warning
Indication range.According to the client of early warning result efficient identification high risk and hospital, and corresponding air control measure is taken (to protect as increased
Take, increase fraud interception rule etc.) reduce Claims Resolution risk.
One embodiment of the invention discloses a kind of method for identifying high risk of fraud comprising obtains Claims Resolution case data, institute
State area, class condition that Claims Resolution case data include multiple customer insureds and corresponding each customer insured and in hospital day
Number；Regional case number of packages, geographic classification condition case number of packages, geographic classification conditional average is calculated according to the Claims Resolution case data to be hospitalized
Number of days and geographic classification condition length of stay standard deviation；Determine high-risk area, wherein the high-risk area describedly
Area's case number of packages is greater than or equal to regional case number of packages threshold value or the accumulative geographic classification condition of the high-risk area is lived
Institute's number of days standard deviation is greater than or equal to accumulative geographic classification condition length of stay standard deviation threshold method；It determines and corresponds to the high risk
The high risk class condition in area sets high risk judgment threshold；Classified according to the high-risk area and the high risk
Condition determines high risk case from the customer insured, wherein the length of stay of the high risk case is greater than or equal to institute
State high risk judgment threshold.
In one embodiment of this invention, the high risk is arranged according to geographic classification conditional average length of stay and judges threshold
Value.
In one embodiment of this invention, the geographic classification condition case number of packages of the high risk class condition be greater than or
Equal to the geographic classification condition length of stay mark of geographic classification condition case number of packages threshold value or the high risk class condition
Quasi- difference is greater than or equal to geographic classification condition length of stay standard deviation threshold method.
In one embodiment of this invention, the class condition is International Classification of Diseases (ICD).
In one embodiment of this invention, according to the case of the regional ICD average hospital days of target ICD, the target ICD
The regional ICD case number of packages of the length of stay of part and the target ICD calculate the regional ICD length of stay standard of the target ICD
Difference.
In one embodiment of this invention, the class condition is hospital.
In one embodiment of this invention, according to ground district hospital average hospital days, the objective hospital of objective hospital
The length of stay of case and the ground district hospital case number of the objective hospital calculate the regional hospital of the objective hospital
Number of days standard deviation.
One embodiment of the invention discloses a kind of device for identifying high risk of fraud comprising: module is obtained, is used to obtain
Claims Resolution case data, the Claims Resolution case data include the area of multiple customer insureds and corresponding each customer insured, classification
Condition and length of stay；Computing module is used to calculate regional case number of packages, geographic classification item according to the Claims Resolution case data
Part case number of packages, geographic classification conditional average length of stay and geographic classification condition length of stay standard deviation；Judgment module is used
In determining that high-risk area and determining corresponds to the high risk class condition of the high-risk area, wherein the high risk
The regional case number of packages in area is greater than or equal to the accumulative area of regional case number of packages threshold value or the high-risk area
Class condition length of stay standard deviation is greater than or equal to accumulative geographic classification condition length of stay standard deviation threshold method；And setting mould
Block is used to set high risk judgment threshold；Wherein, the judgment module is used for according to the high-risk area and the height
Classification of risks condition determines high risk case, wherein the length of stay of the high risk case is greater than from the customer insured
Or it is equal to the high risk judgment threshold.
One embodiment of the invention discloses a kind of electronic equipment comprising: one or more processors；Database, for depositing
Storage Claims Resolution case data；And storage device, for storing one or more programs, when one or more of programs are by described one
When a or multiple processors execute, so that the method that one or more of processors realize the high risk of fraud of identification.
One embodiment of the invention discloses a kind of computer-readable medium, is stored with computer program, the computer journey
The method of the high risk of fraud of identification is realized when sequence is executed by processor.
Technological means of the invention may achieve multiple technical effects.The case data for example, the client based on the past settles a claim, build
Vertical high risk client Early-warning Model and high risk hospital Early-warning Model, the client and doctor of automatic identification high risk and fraud
Institute.End is protected in core by the high risk client that identifies to Early-warning Model, hospital and takes corresponding measure, and can achieve reduces high wind
Dangerous client's insurance risk, it is final to reduce Claims Resolution risk, the effect of net income increase.
Detailed description of the invention
Fig. 1 shows the flow chart of the method for the high risk of fraud of identification of one embodiment of the invention.
Fig. 2 shows the flow chart of the method for the high risk of fraud of identification of another embodiment of the present invention.
Fig. 3 shows the flow chart of the method for the client of the risk of fraud high for identification of another embodiment of the present invention.
Fig. 4 shows the flow chart of the method for the hospital of the risk of fraud high for identification of another embodiment of the present invention.
Fig. 5 shows the schematic diagram of the device of the high risk of fraud of identification of one embodiment of the invention.
Specific embodiment
Spirit for a better understanding of the present invention makees furtherly it below in conjunction with part preferred embodiment of the invention
It is bright.
The embodiment of the present invention is manufactured and used in order to enable to have in the art usually intellectual, is described below
The case where for a special application and its condition.The various modification modes carried out for the embodiment of the present invention, to this
It is obvious for having usually intellectual in technical field.And herein defined in General Principle, without departing from of the invention
Under the spirit and scope of embodiment, it can be used for other embodiments and application.Therefore, embodiments of the present invention are not limited to
In the embodiment having been displayed, and the available most broad range being consistent with the principle in this disclosure with feature.
Fig. 1 shows the flow chart of the method 100 of the high risk of fraud of identification of one embodiment of the invention.Identify high risk of fraud
Method include step 101 to 107.
In a step 101, Claims Resolution case data are obtained.Case data of settling a claim include that multiple customer insureds and correspondence are each
Area, class condition and the length of stay of customer insured.It is province, city, county or other regional scopes that area, which for example can be,.
Class condition for example can be International Classification of Diseases (ICD) or hospital.Case data of settling a claim are, for example, internet insurance company
The past client Claims Resolution data of accumulation.
In a step 102, according to Claims Resolution case data calculate regional case number of packages, geographic classification condition case number of packages, distinguish
Class conditional average length of stay and geographic classification condition length of stay standard deviation.In other words, each area is first calculated to be had
The regional case number of packages having, then according to different class conditions, calculate each area area possessed under different class conditions
Class condition case number of packages and geographic classification conditional average length of stay, then calculate the geographic classification of different class conditions again
Condition length of stay standard deviation.
In step 103, high-risk area is determined.The regional case number of packages of high-risk area is greater than or equal to regional case number of packages
Threshold value or the accumulative geographic classification condition length of stay standard deviation of high-risk area are greater than or equal to accumulative geographic classification item
Part length of stay standard deviation threshold method.In other words, when the regional case number of packages in some area is judged more than or equal to setting
Regional case number of packages threshold value or the accumulative geographic classification condition length of stay standard deviation in some area are judged and are greater than or wait
When the accumulative geographic classification condition length of stay standard deviation threshold method of setting, some area is with will being confirmed as high risk
Area.Wherein, accumulative geographic classification condition length of stay standard deviation refers to that all regions class condition is hospitalized in some area
The summation of number of days standard deviation.It is noted that can determine one or more high-risk areas in step 103.
On the other hand, when some area regional case number of packages be judged less than setting regional case number of packages threshold value, and
The accumulative geographic classification condition length of stay standard deviation in some area is judged the accumulative geographic classification condition less than setting
When length of stay standard deviation threshold method, some area is judged as non-high-risk area, and terminates for some area
Processing routine carries out processing routine to another area again later.
At step 104, the high risk class condition for corresponding to high-risk area is determined.The area of high risk class condition
Class condition case number of packages is greater than or equal to the geographic classification item of geographic classification condition case number of packages threshold value or high risk class condition
Part length of stay standard deviation is greater than or equal to geographic classification condition length of stay standard deviation threshold method.In other words, when some height
The geographic classification condition case number of packages of some class condition of risk region is greater than or equal to geographic classification condition case number of packages threshold value,
Or the geographic classification condition length of stay standard deviation of some class condition of some high-risk area is greater than or equal to ground
When area's class condition length of stay standard deviation threshold method, some class condition of some high-risk area will be confirmed as height
Classification of risks condition.It is noted that can determine one or more high risk class conditions at step 104.
On the other hand, when the geographic classification condition case number of packages of some class condition of some high-risk area is less than ground
The geographic classification condition of area's class condition case number of packages threshold value and some class condition of some high-risk area is hospitalized day
When number standard deviation is less than geographic classification condition length of stay standard deviation threshold method, some class condition of some high-risk area
It is judged as non-high risk class condition, and terminates the processing journey of some class condition for some high-risk area
Sequence carries out processing routine to another class condition again later.If all classification condition of some high-risk area all by
It is judged as non-high risk class condition, then terminates the processing routine for some high-risk area, later again to another height
Risk region carries out processing routine.
In step 105, high risk judgment threshold is set.In some embodiments, it is lived according to geographic classification conditional average
High risk judgment threshold is arranged in institute's number of days.In some embodiments, high risk judgment threshold is calculated using following formula (1):
Geographic classification conditional average length of stay × N (1)
Wherein, N is positive integer, and geographic classification conditional average length of stay calculates in a step 102.
In step 106, high risk case is determined.In some embodiments, according to high-risk area and high risk point
Class condition determines high risk case from customer insured.Wherein, the length of stay of high risk case is sentenced more than or equal to high risk
Disconnected threshold value.In the range of high-risk area and high risk class condition, if the length of stay of some customer insured is big
In or equal to formula (1) calculated high risk judgment threshold, then some customer insured can be confirmed as high risk case.Separately
On the one hand, if the length of stay of some customer insured is less than formula (1) calculated high risk judgment threshold, some
Customer insured is determined as non-high risk case, and terminates the processing routine for being directed to some customer insured, later again to another
One customer insured carries out processing routine.
Internet insurance company can take corresponding air control measure for high risk case after determining high risk case
(such as additional premium increases fraud interception rule) reduces Claims Resolution risk.
It to sum up states, client's Claims Resolution case data based on internet insurance company the past, the present invention can establish high risk
Case Early-warning Model, the case of automatic identification high risk and fraud.Existed by the high risk case identified to Early-warning Model
Core protects end and takes corresponding measure, and can achieve reduces high risk case insurance risk, final to reduce Claims Resolution risk, net income increase
Effect.
Fig. 2 shows the flow chart of the method 200 of the high risk of fraud of identification of another embodiment of the present invention.The high fraud wind of identification
The method of danger includes step 201 to 208.Method 200 and the difference of method 100 are: method 200 further includes step 203 and step
Rapid 205.In addition, the step 201 of method 200,202,204,206,207,208 with the step 101 of method 100,102,103,
104,105,106 is identical, is not repeated to describe below.
In step 203, regional case number of packages threshold value and accumulative geographic classification condition length of stay standard deviation threshold method are set.
Regional case number of packages threshold value and accumulative geographic classification condition length of stay standard deviation threshold method for example can be internet insurance company
According to requiring selected value.
In step 205, geographic classification condition case number of packages threshold value and geographic classification condition length of stay standard deviation are set
Threshold value.Geographic classification condition case number of packages threshold value and geographic classification condition length of stay standard deviation threshold method for example can be internet
Insurance company is according to requiring selected value.
Whereby, internet insurance company can select regional case number of packages threshold value, accumulative geographic classification condition length of stay mark
Quasi- difference threshold value, geographic classification condition case number of packages threshold value and geographic classification condition length of stay standard deviation threshold method, to screen high risk
Area and high risk class condition, to meet the requirement of internet insurance company.
Fig. 3 shows the flow chart of the method 300 of the client of the risk of fraud high for identification of another embodiment of the present invention.With
In the method for the client for identifying high risk of fraud include step 301 to 308.In the present embodiment, class condition is, for example, the world
Classification of diseases (ICD).Claims Resolution case data are, for example, the past client Claims Resolution data of internet insurance company accumulation.
In step 301, Claims Resolution case data are obtained.Case data of settling a claim include that multiple customer insureds and correspondence are each
Area, ICD and the length of stay of customer insured.It is province, city, county or other regional scopes that area, which for example can be,.
In step 302, regional case number of packages, area ICD case number of packages, area ICD is calculated according to Claims Resolution case data to be averaged
Length of stay and area ICD length of stay standard deviation.In other words, area case number of packages possessed by each area is first calculated,
Again according to different ICD, calculates each area area ICD case number of packages possessed at different ICD and area ICD is averagely lived
Institute's number of days then calculates the regional ICD length of stay standard deviation of different ICD again.Wherein, regional ICD average hospital days can
Being calculated using following formula (2):
P is area code, avghidPm,ICDnIt is the regional ICD average hospital days of m-th of area, n-th of ICD,It is the length of stay of m-th of area n-th of ICD, i-th of case, cntPm,ICDnIt is m-th of area, n-th of ICD
Regional ICD case number of packages.
Regional ICD length of stay standard deviation can be to be calculated using following formula (3):
stdhidPm,ICDnIt is the regional ICD length of stay standard deviation of m-th of area, n-th of ICD.
In step 303, regional case number of packages threshold value and accumulative area ICD length of stay standard deviation threshold method are set.Area
Case number of packages threshold value and accumulative area ICD length of stay standard deviation threshold method are, for example, internet insurance company according to requiring to select
Value.
In step 304, high-risk area is determined.The regional case number of packages of high-risk area is greater than or equal to regional case number of packages
Threshold value or the accumulative regional ICD length of stay standard deviation of high-risk area are greater than or equal to accumulative area ICD length of stay
Standard deviation threshold method.In other words, following formula (4), as the regional case number of packages (cnt in some areaPm) be judged and be greater than or wait
In the regional case number of packages threshold value (scnt of settingP),
cntPm≥scntP (4)
Or following formula (5), the accumulative regional ICD length of stay standard deviation (stdhid in some areaPm,ICD)
It is judged the accumulative area ICD length of stay standard deviation threshold method (tstdhid more than or equal to settingP) when,
cntPm,ICDIt is the ICD number in some area, ICDjRepresent j-th of ICD.
Some area will be confirmed as high-risk area.It is noted that can determine in step 304 one or more
A high-risk area.
On the other hand, when some area regional case number of packages be judged less than setting regional case number of packages threshold value, and
The accumulative regional ICD length of stay standard deviation in some area is judged the accumulative area ICD length of stay mark less than setting
When quasi- difference threshold value, some area is judged as non-high-risk area, and terminates the processing routine for some area, it
Processing routine is carried out to another area again afterwards.
In step 305, area ICD case number of packages threshold value and area ICD length of stay standard deviation threshold method are set.Area
ICD case number of packages threshold value and area ICD length of stay standard deviation threshold method are, for example, internet insurance company according to requiring to select
Value.
Within step 306, the high risk ICD for corresponding to high-risk area is determined.The regional ICD case number of packages of high risk ICD
It is greater than or equal to ground more than or equal to regional ICD case number of packages threshold value or the regional ICD length of stay standard deviation of high risk ICD
Area ICD length of stay standard deviation threshold method.In other words, following formula (6), as some ICD of some high-risk area
Regional ICD case number of packages (cntPm,ICDn) it is greater than or equal to area ICD case number of packages threshold value (scntP,ICD),
cntPm,ICDn≥scntP,ICD (6)
Or following formula (7), the regional ICD length of stay standard deviation of some ICD of some high-risk area
(stdhidPm,ICDn) it is greater than or equal to area ICD length of stay standard deviation threshold method (tstdhidP,ICD) when,
stdhidPm,ICDn≥tstdhidP,ICD (7)
Some ICD of some high-risk area will be confirmed as high risk ICD.It is noted that in step 306
In can determine one or more high risks ICD.
On the other hand, when the regional ICD case number of packages of some ICD of some high-risk area is less than area ICD case
The regional ICD length of stay standard deviation of some ICD of number threshold value and some high-risk area is less than area ICD and is hospitalized
When number of days standard deviation threshold method, some ICD of some high-risk area is judged as non-high risk ICD, and terminates to be directed to
The processing routine of some ICD of some high-risk area carries out processing routine to another ICD again later.If a certain
Whole ICD of a high-risk area are judged as non-high risk ICD, then terminate the processing journey for some high-risk area
Sequence carries out processing routine to another high-risk area again later.
In step 307, high risk judgment threshold is set.High risk judgment threshold can be to be counted using following formula (8)
It calculates:
avghidPm,ICDn×N (8)
Wherein, N is positive integer, regional ICD average hospital days (avghidPm,ICDn) calculated in step 302.
In step 308, high risk client is determined.In some embodiments, according to high-risk area and high risk
ICD determines high risk client from customer insured.Wherein, the length of stay of high risk client judges more than or equal to high risk
Threshold value.In the range of high-risk area and high risk ICD, following formula (9), if some customer insured's is hospitalized
Number of daysMore than or equal to the calculated high risk judgment threshold of formula (8),
N is positive integer, represents the threshold multiple of setting.
Then some customer insured can be confirmed as high risk client.On the other hand, if some customer insured's lives
Institute's number of days is less than formula (8) calculated high risk judgment threshold, then some customer insured is determined as non-high risk visitor
Family, and terminate the processing routine for being directed to some customer insured, processing routine is carried out to another customer insured again later.
Internet insurance company can take corresponding air control measure for high risk client after determining high risk client
(such as additional premium increases fraud interception rule) reduces Claims Resolution risk.
It to sum up states, client's Claims Resolution case data based on internet insurance company the past, the present invention can establish high risk
Client's Early-warning Model, the client of automatic identification high risk and fraud.By existing to the high risk client that Early-warning Model identifies
Core protects end and takes corresponding measure, and can achieve reduces high risk client insurance risk, final to reduce Claims Resolution risk, net income increase
Effect.
Fig. 4 shows the flow chart of the method 400 of the hospital of the risk of fraud high for identification of one embodiment of the invention.For
The method for identifying the hospital of high risk of fraud includes step 401 to 408.In the present embodiment, class condition is, for example, hospital.
In step 401, Claims Resolution case data are obtained.Case data of settling a claim include that multiple customer insureds and correspondence are each
Area, hospital and the length of stay of customer insured.Claims Resolution case data are, for example, the past doctor of internet insurance company accumulation
Institute's Claims Resolution data.It is province, city, county or other regional scopes that area, which for example can be,.
In step 402, according to Claims Resolution case data calculate regional case number of packages, district hospital's case number, district hospital it is flat
Equal length of stay and regional hospital number of days standard deviation.In other words, area case possessed by each area is first calculated
Number, then according to different hospitals, calculate each area possessed ground district hospital case number and area doctor under different hospitals
Institute's average hospital days then calculate the regional hospital number of days standard deviation of different hospitals again.Wherein, ground district hospital is average
Length of stay can be to be calculated using following formula (10):
P is area code, avghidPm,HSPnIt is the ground district hospital average hospital days of m-th of regional n-th of hospital,It is the length of stay of m-th of regional n-th of hospital, i-th of case, cntPm,HSPnIt is n-th of m-th of area doctor
The ground district hospital case number of institute.
Regional hospital number of days standard deviation can be to be calculated using following formula (11):
stdhidPm,HSPnIt is the regional hospital number of days standard deviation of m-th of regional n-th of hospital.
In step 403, regional case number of packages threshold value and cumulatively district hospital's length of stay standard deviation threshold method are set.Area
Case number of packages threshold value and cumulatively district hospital's length of stay standard deviation threshold method are, for example, that internet insurance company is selected according to requiring
Fixed value.
In step 404, high-risk area is determined.The regional case number of packages of high-risk area is greater than or equal to regional case number of packages
Threshold value or the accumulative regional hospital number of days standard deviation of high-risk area are greater than or equal to cumulatively district hospital and are hospitalized day
Number standard deviation threshold method.In other words, following formula (12), as the regional case number of packages (cnt in some areaPm) be judged be greater than or
Equal to the regional case number of packages threshold value (scnt of settingP),
cntPm≥scntP (12)
Or following formula (13), the accumulative regional hospital number of days standard deviation in some area
It is judged the cumulatively district hospital length of stay standard deviation threshold method (tstdhid more than or equal to settingP) when,
cntP,HSPIt is hospital's number in some area, HSPjRepresent j-th of hospital.
Some area will be confirmed as high-risk area.It is noted that can determine in step 404 one or more
A high-risk area.
On the other hand, when some area regional case number of packages be judged less than setting regional case number of packages threshold value, and
The accumulative regional hospital number of days standard deviation in some area is judged cumulatively district hospital's length of stay less than setting
When standard deviation threshold method, some area is judged as non-high-risk area, and terminates the processing routine for some area,
Processing routine is carried out to another area again later.
In step 405, setting ground district hospital's case number threshold value and regional hospital number of days standard deviation threshold method.Area
Hospital's case number of packages threshold value and regional hospital number of days standard deviation threshold method are, for example, that internet insurance company is selected according to requiring
Fixed value.
In a step 406, the high risk hospital for corresponding to high-risk area is determined.The ground district hospital case of high risk hospital
Number is greater than or equal to ground district hospital's case number threshold value or the regional hospital number of days standard deviation of high risk hospital is greater than or waits
In regional hospital number of days standard deviation threshold method.In other words, following formula (14), when some of some high-risk area
The ground district hospital case number (cnt of hospitalPm,HSPn) it is greater than or equal to ground district hospital case number threshold value (scntP,HSP),
cntPm,HSPn≥scntP,HSP (14)
Or following formula (15), the regional hospital number of days standard deviation of some hospital of some high-risk area
(stdhidPm,HSPn) it is greater than or equal to regional hospital number of days standard deviation threshold method (tstdhidP,HSP) when,
stdhidPm,HSPn≥tstdhidP,HSP (15)
Some hospital of some high-risk area will be confirmed as high risk hospital.It is noted that in step
One or more high risk hospitals can be determined in 406.
On the other hand, when the ground district hospital case number of some hospital of some high-risk area is less than ground district hospital case
The regional hospital number of days standard deviation of some hospital of number of packages threshold value and some high-risk area is less than ground district hospital
When length of stay standard deviation threshold method, some hospital of some high-risk area is judged as non-high risk hospital, and ties
Beam carries out processing routine to another hospital again later for the processing routine of some hospital of some high-risk area.
If whole hospitals of some high-risk area are all judged as non-high risk hospital, terminate for some high risk
The processing routine in area carries out processing routine to another high-risk area again later.
In step 407, high risk judgment threshold is set.High risk judgment threshold can be to be counted using following formula (16)
It calculates:
avghidPm,HSPn×N (16)
Wherein, N is positive integer, ground district hospital average hospital days (avghidPm,HSPn) calculated in step 402.
In a step 408, high risk hospital is determined.In some embodiments, it is cured according to high-risk area and high risk
Institute determines high risk case from customer insured.Wherein, the length of stay of high risk case judges threshold more than or equal to high risk
Value.In the range of high-risk area and high risk hospital, following formula (17), if some is insured, hospital is hospitalized
Number of daysMore than or equal to the calculated high risk judgment threshold of formula (16),
Then some hospital of insuring can be confirmed as high risk hospital.On the other hand, if some is insured hospital live
Institute's number of days is less than formula (16) calculated high risk judgment threshold, then some hospital of insuring is determined as non-high risk doctor
Institute, and terminate the processing routine for being directed to some hospital of insuring, processing routine is carried out to another hospital of insuring again later.
Internet insurance company can take corresponding air control measure for high risk hospital after determining high risk hospital
(such as additional premium increases fraud interception rule) reduces Claims Resolution risk.
It to sum up states, client's Claims Resolution case data based on internet insurance company the past, the present invention can establish high risk
Hospital's Early-warning Model, the hospital of automatic identification high risk and fraud.Existed by the high risk hospital identified to Early-warning Model
Core protects end and takes corresponding measure, and can achieve reduces high risk hospital insurance risk, final to reduce Claims Resolution risk, net income increase
Effect.
Fig. 5 shows the schematic diagram of the device 500 of the high risk of fraud of identification of one embodiment of the invention.Identify high risk of fraud
Device include obtaining module 501, computing module 502, judgment module 503 and setting module 504.
Module 501 is obtained for obtaining Claims Resolution case data.Case data of settling a claim include multiple customer insureds and correspondence
Area, class condition and the length of stay of each customer insured.Class condition for example can be International Classification of Diseases (ICD) or
Person hospital.
Computing module 502 be used to be calculated according to the Claims Resolution case data area case number of packages, geographic classification condition case number of packages,
Geographic classification conditional average length of stay and geographic classification condition length of stay standard deviation.
Judgment module 503 is used to determine high-risk area and determines the high risk classification item for corresponding to high-risk area
Part.Wherein, the regional case number of packages of high-risk area is greater than or equal to the accumulative of regional case number of packages threshold value or high-risk area
Geographic classification condition length of stay standard deviation is greater than or equal to accumulative geographic classification condition length of stay standard deviation threshold method.
Setting module 504 is for setting high risk judgment threshold.Wherein, judgment module 503 is used for according to high-risk area
And high risk class condition, from customer insured determine high risk case, the length of stay of medium or high risk case be greater than or
Equal to the high risk judgment threshold.
It identifies that the device 500 of high risk of fraud can carry out as shown in Figure 1, Figure 2, Fig. 3 and Fig. 4 method, is not repeated herein
Narration.
In one embodiment, the method for the high risk of fraud of identification as above can be implemented in electronic equipment.Electronics is set
Standby includes: one or more processors；Database, for storing Claims Resolution case data；And storage device, for store one or
Multiple programs, when one or more of programs are executed by one or more of processors, so that one or more of
Processor obtains Claims Resolution case data and realizes the method for identifying high risk of fraud.
In one embodiment, the method for the high risk of fraud of identification as above can be implemented in computer-readable medium.
Computer-readable medium storage has computer program, and the side for identifying high risk of fraud is realized when computer program is executed by processor
Method.
Technology contents and technical characterstic of the invention have revealed that as above, however those skilled in the art still may base
Make various replacements and modification without departing substantially from spirit of that invention in teachings of the present invention and announcement.Therefore, protection model of the invention
The revealed content of embodiment should be not limited to by enclosing, and should include various without departing substantially from replacement and modification of the invention, and be this patent
Application claims are covered.
Claims (10)
1. a kind of method for identifying high risk of fraud characterized by comprising
Claims Resolution case data are obtained, the Claims Resolution case data include the ground of multiple customer insureds and corresponding each customer insured
Area, class condition and length of stay；
Regional case number of packages, geographic classification condition case number of packages, geographic classification conditional average is calculated according to the Claims Resolution case data to live
Institute's number of days and geographic classification condition length of stay standard deviation；
High-risk area is determined, wherein the regional case number of packages of the high-risk area is greater than or equal to regional case number of packages threshold
Value or the accumulative geographic classification condition length of stay standard deviation of the high-risk area are greater than or equal to accumulative area
Class condition length of stay standard deviation threshold method；
Determine the high risk class condition for corresponding to the high-risk area；
Set high risk judgment threshold；And
According to the high-risk area and the high risk class condition, high risk case is determined from the customer insured,
Wherein the length of stay of the high risk case is greater than or equal to the high risk judgment threshold.
2. the method according to claim 1, wherein being arranged according to the geographic classification conditional average length of stay
The high risk judgment threshold.
3. the method according to claim 1, wherein the wherein geographic classification of the high risk class condition
Condition case number of packages is greater than or equal to distinguishing describedly for geographic classification condition case number of packages threshold value or the high risk class condition
Class condition length of stay standard deviation is greater than or equal to geographic classification condition length of stay standard deviation threshold method.
4. the method according to claim 1, wherein wherein the class condition is International Classification of Diseases ICD.
5. according to the method described in claim 4, it is characterized in that, wherein according to the regional ICD average time in hospital day of target ICD
The length of stay of the case of several, the described target ICD and the regional ICD case number of packages of the target ICD calculate the ground of the target ICD
Area ICD length of stay standard deviation.
6. the method according to claim 1, wherein wherein the class condition is hospital.
7. according to the method described in claim 6, it is characterized in that, wherein according to the ground district hospital average time in hospital day of objective hospital
The length of stay of the case of several, the described objective hospital and the ground district hospital case number of the objective hospital calculate the objective hospital
Regional hospital number of days standard deviation.
8. a kind of device for identifying high risk of fraud, characterized in that include:
Acquisition module, for obtaining Claims Resolution case data, the Claims Resolution case data include multiple customer insureds and correspond to every
Area, class condition and the length of stay of one customer insured；
Computing module is used to calculate regional case number of packages, geographic classification condition case number of packages, area according to the Claims Resolution case data
Class condition average hospital days and geographic classification condition length of stay standard deviation；
Judgment module is used to determine high-risk area and determines the high risk classification item for corresponding to the high-risk area
Part, wherein the regional case number of packages of the high-risk area is greater than or equal to regional case number of packages threshold value or the high risk
The accumulative geographic classification condition length of stay standard deviation in area is greater than or equal to accumulative geographic classification condition length of stay
Standard deviation threshold method；And
Setting module is used to set high risk judgment threshold；
Wherein, the judgment module is used to be insured according to the high-risk area and the high risk class condition from described
High risk case is determined in client, wherein the length of stay of the high risk case, which is greater than or equal to the high risk, judges threshold
Value.
9. a kind of electronic equipment characterized by comprising
One or more processors；
Database, for storing Claims Resolution case data；And
Storage device, for storing one or more programs, when one or more of programs are by one or more of processing
When device executes, so that one or more of processors realize that the identification height as described in any one of claims 1 to 7 cheats wind
The method of danger.
10. a kind of computer-readable medium, is stored with computer program, which is characterized in that the computer program is processed
The method of the high risk of fraud of identification as described in any one of claims 1 to 7 is realized when device executes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910105306.6A CN109785174A (en) | 2019-02-01 | 2019-02-01 | The method and apparatus for identifying high risk of fraud |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910105306.6A CN109785174A (en) | 2019-02-01 | 2019-02-01 | The method and apparatus for identifying high risk of fraud |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109785174A true CN109785174A (en) | 2019-05-21 |
Family
ID=66503964
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910105306.6A Pending CN109785174A (en) | 2019-02-01 | 2019-02-01 | The method and apparatus for identifying high risk of fraud |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109785174A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114334043A (en) * | 2021-12-30 | 2022-04-12 | 上海柯林布瑞信息技术有限公司 | Medical insurance-based diagnosis and treatment critical path monitoring method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104408547A (en) * | 2014-10-30 | 2015-03-11 | 浙江网新恒天软件有限公司 | Data-mining-based detection method for medical insurance fraud behavior |
| CN107203944A (en) * | 2017-05-22 | 2017-09-26 | 中国平安人寿保险股份有限公司 | Visualize data monitoring method and device |
| CN107609980A (en) * | 2017-09-07 | 2018-01-19 | 平安医疗健康管理股份有限公司 | Medical data processing method, device, computer equipment and storage medium |
-
2019
- 2019-02-01 CN CN201910105306.6A patent/CN109785174A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104408547A (en) * | 2014-10-30 | 2015-03-11 | 浙江网新恒天软件有限公司 | Data-mining-based detection method for medical insurance fraud behavior |
| CN107203944A (en) * | 2017-05-22 | 2017-09-26 | 中国平安人寿保险股份有限公司 | Visualize data monitoring method and device |
| CN107609980A (en) * | 2017-09-07 | 2018-01-19 | 平安医疗健康管理股份有限公司 | Medical data processing method, device, computer equipment and storage medium |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114334043A (en) * | 2021-12-30 | 2022-04-12 | 上海柯林布瑞信息技术有限公司 | Medical insurance-based diagnosis and treatment critical path monitoring method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mensah et al. | Aboveground biomass and carbon in a South African mistbelt forest and the relationships with tree species diversity and forest structures | |
| CN107910068A (en) | Insure health risk Forecasting Methodology, device, equipment and the storage medium of user | |
| CN104657705B (en) | Pattern recognition device and data entry method towards pattern recognition device | |
| CN108876636B (en) | Intelligent air control method, system, computer equipment and storage medium for claim settlement | |
| CN110610431B (en) | Intelligent claim settlement method and intelligent claim settlement system based on big data | |
| CN113270197A (en) | Health prediction method, system and storage medium based on artificial intelligence | |
| WO2019041518A1 (en) | Electronic device, method and system for examination of insurance claim, and computer-readable storage medium | |
| EP1561178A2 (en) | Fraud and abuse detection and entity profiling in hierarchical coded payment systems | |
| Shi et al. | Amur tiger stripes: Individual identification based on deep convolutional neural network | |
| CN106408423A (en) | Method and system for risk assessment and method for constructing system for risk assessment | |
| Martínez et al. | Disentangling habitat use by frugivorous birds: constant interactive effects of forest cover and fruit availability | |
| CN105809343A (en) | Underwriting and actuarial database system used for evaluating insurance subject risks | |
| CN109360109A (en) | Intelligent Claims Review householder method and system based on big data analysis and processing | |
| CN110321350B (en) | Survival authentication method and system based on data restoration and active learning verification | |
| CN113610934B (en) | Image brightness adjustment method, device, equipment and storage medium | |
| CN116681923A (en) | Automatic ophthalmic disease classification method and system based on artificial intelligence | |
| CN109785174A (en) | The method and apparatus for identifying high risk of fraud | |
| CN109359802B (en) | Inverse selection risk determination method, system, computer device and storage medium | |
| CN110706121A (en) | Method and device for determining medical insurance fraud result, electronic equipment and storage medium | |
| CN116350203B (en) | Physical testing data processing method and system | |
| CN109190489A (en) | A kind of abnormal face detecting method based on reparation autocoder residual error | |
| Webb et al. | Developing spatially‐explicit weighting factors to account for bias associated with missed GPS fixes in resource selection studies | |
| CN117372173A (en) | Agricultural risk processing method, device, equipment and storage medium based on image recognition | |
| CN116844730A (en) | Modeling method and system of cognitive disorder recognition model | |
| CN110210985A (en) | Basic medical insurance policies adjust the quantization method influenced on fund expenditure and system |
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 |
Application publication date: 20190521 |
|
| RJ01 | Rejection of invention patent application after publication |