Automated method for identification of reinsurance claims
First Claim
1. A method for identifying insurance claims having a relationship with reinsurance related catastrophic events, comprising the steps of:
- providing a catastrophic database having identified cluster nodes of catastrophic events derived by reported parameters, each said cluster node having an earth position parameter, and a date of occurrence parameter;
providing an insurance claims database of claim reports, each such claim report thereof having, an earth location datum and a date datum;
deriving a place degree of belonging for each said earth location datum of each said claim report with respect to each said cluster node earth position parameter;
deriving a date degree of belonging for each said date datum of each said claims report with respect to each said date of occurrence parameter;
selecting a said identified cluster node for each said claim report as that best cluster node optimally corresponding with the derived place degree of belonging and date degree of belonging; and
generating a tabular claims file compilation incorporating data from each said claim report, each said claim report being associated with said selected identified best cluster node, with said best place degree of belonging and date degree of belonging.
1 Assignment
0 Petitions
Accused Products
Abstract
An automated method for identifying those claims of a raw claims insurance claims database for which reinsurance is applicable. The method develops a database of uniquely clustered catastrophic events such as storm reports. This "CatNodes" database may be developed manually or automatically through the use of a neural network approach. A fuzzy degree of belonging is employed to quantify the likelihood that a given insurance claim is properly associated with a given catastrophic event or storm. The assignment of a degree of belonging is derived by approaches which consider four factors: the date of the loss; the location of the loss; the type of the loss; and the presence of special keywords in the claim description.
258 Citations
39 Claims
-
1. A method for identifying insurance claims having a relationship with reinsurance related catastrophic events, comprising the steps of:
-
providing a catastrophic database having identified cluster nodes of catastrophic events derived by reported parameters, each said cluster node having an earth position parameter, and a date of occurrence parameter; providing an insurance claims database of claim reports, each such claim report thereof having, an earth location datum and a date datum; deriving a place degree of belonging for each said earth location datum of each said claim report with respect to each said cluster node earth position parameter; deriving a date degree of belonging for each said date datum of each said claims report with respect to each said date of occurrence parameter; selecting a said identified cluster node for each said claim report as that best cluster node optimally corresponding with the derived place degree of belonging and date degree of belonging; and generating a tabular claims file compilation incorporating data from each said claim report, each said claim report being associated with said selected identified best cluster node, with said best place degree of belonging and date degree of belonging.
-
-
2. A method for producing, from a memory retained group of insurance claim reports, a compilation of insurance claims having a relationship with catastrophes for which reinsurance coverage is applicable, comprising the steps of:
-
providing a city database wherein a plurality of cities name entries are retained in memory, each being retained in combination with geographic coordinates identifying their position on earth; opening a claims database having fields including a loss city field, a loss state field, a claim identification field, and a loss date field; importing claim data available from each said claim report of said memory retained group of insurance claims reports into corresponding said fields; for each said claim report, scanning said city database to identify a city name match and retaining said geographic coordinates as claim geographic coordinates in memory in conjunction with said claims report when a city name match occurs; providing a catastrophic database having identified clusters of catastrophic events derived by reported parameters, each said cluster having an earth position parameter and a date of occurrence parameter; deriving a degree of belonging corresponding with each said claim geographic coordinates with respect to each said cluster earth position parameter, and corresponding with each said loss date of each said claim report with respect to each said cluster date of occurrence parameter; selecting the best cluster representing a best said degree of belonging for each said claim report; and generating a tabular claims file compilation incorporating data from each said claim report, each said claims report tabulated therein being combined with said selected best cluster of catastrophic events. - View Dependent Claims (3, 4, 5, 6, 7, 8, 9, 10)
-
-
11. A method for identifying insurance claim having a relationship with reinsurance related catastrophic events, comprising the steps of:
-
(a1) providing a catastrophic database as a memory retained compilation of catastrophic event records, each such record having an event identification, a date of event occurrence, and a catastrophic geographic coordinate identifying its position on earth; (a2) selecting a first place vigilance value; (a3) selecting a first date vigilance value; (a4) opening a catnodes database; (a5) selecting an initial node as one said event record with a node coordinate and a node date corresponding respectively with said catastrophic geographic coordinates and said date of event occurrence; (a6) scanning said catastrophic database and for each record thereof, deriving a place degree of belonging in correspondence with said node coordinate and said catastrophic geographic coordinate, and further deriving a date degree of belonging in correspondence with said node date and said date of event occurrence; (a7) comparing the value of said place degree of belonging with the said first place vigilance value and diminishing said place degree of belonging when the value thereof is less than said first place vigilance value; (a8) comparing the value of said date degree of belonging with said first date vigilance value and diminishing said date degree of belonging when the value thereof is less than said first date vigilance value; (a9) determining an overall degree of belonging in correspondence with said place degree of belonging and said date degree of belonging; (a10) determining a best degree of belonging from all overall degree of belonging values derived from said scanning step (a6) and assigning the place degree of belonging and the date degree of belonging corresponding with said best degree of belonging respectively as best place degree of belonging and best date degree of belonging; (a11) creating a new node corresponding with that event record deriving said best degree of belonging when the value of said best place degree of belonging is less than said first place vigilance value or when the value of said best date degree of belonging is less than said first date vigilance value; (a12) entering said new node into said catnodes database; (a13) adjusting said node coordinates and said node date in correspondence with said best place degree of belonging and best date degree of belonging when the value of said best place degree of belonging is greater than said first place vigilance value and said best date degree of belonging is greater than said first date vigilance value; (a14) reiterating steps (a6) through (a13) for each new node created to form a catnodes database; (b) providing an insurance claims database of claim reports, each such claim report thereof having claim place geographic coordinates identifying a position on earth and a claim loss date; (c) deriving a claim place degree of belonging corresponding with each of said claim place geographic coordinates with respect to each said node coordinate of said catnodes database; (d) deriving a claims date degree of belonging corresponding with each said claim loss date with respect to each said node date of said catnodes database; (e) deriving a claims overall degree of belonging corresponding with each said claims place degree of belonging and said claim loss date degree of belonging; (f) selecting the best node of said catnodes database representing a best overall degree of belonging for each said claim report; and generating a tabular ClaimsFile compilation incorporating data from each said claim report, each said claim report tabulated therein being combined with said selected best node. - View Dependent Claims (12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32)
-
-
33. A method for identifying insurance claims having a relationship with reinsurance related catastrophic events, comprising the steps of:
-
providing a catastrophic database having identified cluster nodes of catastrophic events derived by reported parameters, each said cluster node having an earth position parameter, and a date of occurrence parameter; providing an insurance claims database of claim reports, each such claim report thereof having an earth location datum and a date datum; deriving a place degree of belonging for each said earth location datum of each said claim report with respect to each said cluster node earth position parameter; deriving a date degree of belonging for each said date datum of each said claims report with respect to each said date of occurrence parameter; selecting a said identified cluster node for each said claim report as that best cluster node optimally corresponding with the derived place degree of belonging and date degree of belonging by; selectively combining the values of said place degree of belonging and said date degree of belonging to derive an overall degree of belonging, and selecting said optimally corresponding best cluster node as that cluster node corresponding with the best said overall degree of belonging; and generating a tabular claims file compilation incorporating data from each said claim report, each said claim report being associated with said selected identified best cluster node, with said best place degree of belonging and date degree of belonging. - View Dependent Claims (34, 35)
-
-
36. A method for identifying insurance claims having a relationship with reinsurance related catastrophic events, comprising the steps of:
-
providing a catastrophic database having identified cluster nodes of catastrophic events derived by reported parameters, each said cluster node having an earth position parameter, and a date of occurrence parameter; providing an insurance claims database of claim reports, each such claim report thereof having an earth location datum and a date datum; deriving a place degree of belonging for each said earth location datum of each said claim report with respect to each said cluster node earth position parameter; deriving a date degree of belonging for each said date datum of each said claims report with respect to each said date of occurrence parameter; selecting a said identified cluster node for each said claim report as that best cluster node optimally corresponding with the derived place degree of belonging and date degree of belonging; generating a tabular claims file compilation incorporating data from each said claim report, each said claim report being associated with said selected identified best cluster node, with said best place degree of belonging and date degree of belonging; selecting a target identified cluster node; and compiling a tabulation of all said claim records wherein said target identified cluster node is said best cluster node.
-
-
37. A method for identifying insurance claims having a relationship with reinsurance related catastrophic events, comprising the steps of:
-
providing a catastrophic database having identified cluster nodes of catastrophic events derived by reported parameters, each said cluster node having an earth position parameter, and a date of occurrence parameter; providing an insurance claims database of claim reports, each such claim report thereof having an earth location datum and a date datum; deriving a place degree of belonging for each said earth location datum of each said claim report with respect to each said cluster node earth position parameter; deriving a date degree of belonging for each said date datum of each said claims report with respect to each said date of occurrence parameter; selecting a said identified cluster node for each said claim report as that best cluster node optimally corresponding with the derived place degree of belonging and date degree of belonging; generating a tabular claims file compilation incorporating data from each said claim report, each said claim report being associated with said selected identified best cluster node, with said best place degree of belonging and date degree of belonging; and wherein; said catastrophic database earth position parameter includes a fuzzy set with principal geographical coordinates and a territorial domain extending a predetermined distance therefrom; and said place degree of belonging is derived in correspondence with the distance between said principal geographical coordinates and said earth location datum, divided by a fuzzy set territorial domain extending a predetermined distance from said geographical coordinates.
-
-
38. A method for identifying insurance claims having a relationship with reinsurance related catastrophic events, comprising the steps of:
-
providing a catastrophic database having identified cluster nodes of catastrophic events derived by reported parameters, each said cluster node having an earth position parameter, and a date of occurrence parameter; providing an insurance claims database of claim reports, each such claim report thereof having an earth location datum and a date datum; deriving a place degree of belonging for each said earth location datum of each said claim report with respect to each said cluster node earth position parameter; deriving a date degree of belonging for each said date datum of each said claims report with respect to each said date of occurrence parameter, said date degree of belonging being derived in correspondence with the time difference between said claim report date datum and said date of occurrence parameter divided by a predetermined fuzzy set time domain value; selecting a said identified cluster node for each said claim report as that best cluster node optimally corresponding with the derived place degree of belonging and date degree of belonging; and generating a tabular claims file compilation incorporating data from each said claim report, each said claim report being associated with said selected identified best cluster node, with said best place degree of belonging and date degree of belonging. - View Dependent Claims (39)
-
Specification