Outlier system for grouping of characteristics

1 . A device for automatically grouping data based on characteristics, the device comprising:

a communication interface configured to receive data from a plurality of different data sources;

a storage medium configured to store the received data;

a processor capable of operating cooperatively with the communication interface to request data and with the storage medium to query stored data, the processor configured to:

correlate selected parameters within the received data, wherein the correlated selected parameters include at least relationships with other data; and

model the received data by analyzing the received data using the correlated selected parameters to identify attributes associated with particular groups of data using the correlated selected parameters and based on groupings of attributes and relative separation of an attribute grouping from typical groupings of those attributes to identify data points having common attributes; and

the communication interface further configured to:

input the correlated selected parameters and the identified data points having common attributes as additional data sources of the plurality of data sources; and

output results of the modeling including the correlations and the identified data points to at least one system, wherein the outputted correlations and the identified data points improve a process based on groupings of attributes and relative separation of an attribute grouping from typical groupings of those attributes of the at least one system, wherein the improvement in the process includes at least one of an underwriting decision, a claim to be investigated for fraud and claims result.

2 . The device of claim 1 , wherein the model includes plotting the received data in a multivariate form to identify groupings of data points.

3 . The device of claim 1 , wherein the model identifies and derives parameters within the received data according to a type of data.

4 . The device of claim 1 , wherein the received data is related to a loss or a claim in an insurance industry, wherein the loss or the claim includes a loss state, a claimant age, an injury type, or a type of reporting of an injury, wherein the loss or the claim includes physician billing or treatment patterns.

5 . The device of claim 1 , wherein the received data includes a benchmark comparison and identifies a difference from a benchmark.

6 . The device of claim 1 , wherein the model includes predictor variables, wherein the predictor variables include source systems, text mined data, and outlier data.

7 . The device of claim 1 , wherein the model is generated on the received data based on inclusion in or exclusion from a group, wherein the model is performed on the received data with identified attributes associated with a particular group.

8 . A computerized method for automatically grouping data based on characteristics, the method comprising:

receiving, by a processor via a communication interface, data from a plurality of different data sources;

storing, by the processor, the received data in a storage medium;

correlating, by the processor cooperatively operative with the communication interface to request data and with the storage medium to query shared data, selected parameters within the received data, wherein the correlated selected parameters include at least relationships with other data;

modeling, by the processor, the received data by analyzing the received data using the correlated selected parameters to identify attributes associated with particular groups of data using the correlated selected parameters and based on groupings of attributes and relative separation of an attribute grouping from typical groupings of those attributes to identify data points having common attributes;

inputting, by the processor via the communication interface, the correlated selected parameters and the identified data points having common attributes as additional data sources of the plurality of data sources; and

outputting, by the processor, results of the modeling including the correlations and the identified data points to at least one system, wherein the outputted correlations and the identified data points improve a process of the at least one system based on groupings of attributes and relative separation of an attribute grouping from typical groupings of those attributes, wherein the improvement in the process includes at least one of an underwriting decision, a claim to be investigated for fraud and claims result.

9 . The method of claim 8 , wherein the model includes plotting the received data in a multivariate form to identify groupings of data points.

10 . The method of claim 8 , wherein the model identifies and derives parameters within the received data according to a type of data.

11 . The method of claim 8 , wherein the received data is related to a loss or a claim in an insurance industry, wherein the loss or the claim includes a loss state, a claimant age, an injury type, or a type of reporting of an injury, wherein the loss or the claim includes physician billing or treatment patterns.

12 . The method of claim 8 , wherein the received data includes a benchmark comparison and identifies a difference from a benchmark.

13 . The method of claim 8 , wherein the model includes predictor variables, wherein the predictor variables include source systems, text mined data, and outlier data.

14 . The method of claim 8 , wherein the model is generated on the received data based on inclusion in or exclusion from a group, wherein the model is performed on the received data with identified attributes associated with a particular group.

15 . A non-transitory computer readable medium having stored thereon instructions that, when executed by a processor, cause the processor to perform operations comprising:

receiving data from a plurality of different data sources;

storing the received data in a storage medium;

correlating selected parameters within the received data, wherein the correlated selected parameters include at least relationships with other data;

modeling the received data by analyzing the received data using the correlated selected parameters to identify attributes associated with particular groups of data using the correlated selected parameters and based on groupings of attributes and relative separation of an attribute grouping from typical groupings of those attributes to identify data points having common attributes

inputting the correlated selected parameters and the identified data points having common attributes as additional data sources of the plurality of data sources; and

outputting results of the modeling including the correlations and the identified data points to at least one system, wherein the outputted correlations and the identified data points improve a process of the at least one system based on groupings of attributes and relative separation of an attribute grouping from typical groupings of those attributes, wherein the improvement in the process includes at least one of an underwriting decision, a claim to be investigated for fraud and claims result.

16 . The non-transitory computer readable medium of claim 15 , wherein the model includes plotting the received data in a multivariate form to identify groupings of data points.

17 . The non-transitory computer readable medium of claim 15 , wherein the model identifies and derives parameters within the received data according to a type of data.

18 . The non-transitory computer readable medium of claim 15 , wherein the received data is related to a loss or a claim in an insurance industry, wherein the loss or the claim includes a loss state, a claimant age, an injury type, a type of reporting of an injury, physician billing, or treatment patterns.

19 . The non-transitory computer readable medium of claim 15 , wherein the received data includes a benchmark comparison and identifies a difference from a benchmark.

20 . The non-transitory computer readable medium of claim 15 , wherein the model includes predictor variables, wherein the predictor variables include source systems, text mined data, and outlier data.

21 . The non-transitory computer readable medium of claim 15 , wherein the model is generated on the received data based on inclusion in or exclusion from a group, and wherein the modeling is performed on the received data with identified attributes associated with a particular group.