FRISS DETECTION

The Friss Detection module can be accessed in more than one place during the processes. These are:
  • during an apply for a new insurance
  • during a request for a change to an insurance
  • during the claim intake process
  • during or after saving additional details about the claim, for example on the basis of an expert report
  • just before paying out a claim
These are examples within the claim process. Friss Detection can also be used in other lines of business such as during the offer process for mortgages or a claim made from income insurances. This check consists of evaluating the validated fraud indicators. A relevance can be given with a so-called precondition to each indicator and measurement. There is a large difference between indicators that are applicable to a life insurance during the underwriting stage and indicators that must be checked for a claim for vehicle damage. This can be allowed for by using preconditions. The indicators are measured using complex decision tables (knowledge rules). More than one measurement can be defined for each indicator. The value is determined by evaluating using a decision table. Use is made of various internal and external sources of data for the evaluation. The value is multiplied by the weighting factor   of the indicator in order to provide the score for that indicator. The total score of all indicators gives the end result. This is used to decide whether a claim or a new request, for example, should be filtered out of the primary process for separate treatment. The detection can be used in real time or in a batch in the primary process of an insurer. The real time variant can be accessed immediately during the intake of a claim or a new application. The application requesting the access and with this the user of that application is then given an immediate result for the fraud risk for the particular case on the basis of the indicators that are relevant for that case. All results of Friss Detection are saved in the Friss Fraud Database. Reports can be run from this via the Dashboard module. Cases that do not lead to signalling are also registered. The reason for this is naturally to learn from these through reports and analyses.

Allsecur predicts defaulters with Friss

read more

friss animations