Insurance anti-fraud method for identifying drunk driving insurance switching based on operator data

Abstract

The invention discloses an insurance anti-fraud method for identifying drunk driving insurance switching based on operator data, and relates to the technical field of vehicle insurance claim anti-fraud automatic discrimination algorithms. The method comprises four parts, namely an insurance company, an operator PAAS platform (Platform-as-a-Service), an operator deep analysis cloud system and an insurance science and technology company, and each part establishes a TCP (transmission control protocol) connection through an application programming interface, and transmits the related data. The insurance anti-fraud method for identifying the drunk driving insurance switching behaviors adopts a big data modeling technology, establishes a vehicle insurance claim settlement anti-fraud system basedon the operator data, and during the claim settlement case checking process, can mine and analyze from different angles based on the user behavior data, such as the user call records, the call positions, etc., and scores the insurance reports reasonably in a targeted manner.

Description

technical field [0001] The invention relates to the technical field of auto insurance claim anti-fraud automatic discrimination algorithm, in particular to an insurance anti-fraud method based on operator data identification of drunk driving insurance and subcontracting. Background technique [0002] At present, there are a lot of fraudulent behaviors in the field of auto insurance claims. The phenomenon of drunk driving and subcontracting is relatively common in auto insurance fraud cases. After a driver has an accident due to drunk driving, in order to avoid accident responsibility and defraud insurance compensation, he will ask his friends to pretend to be the driver to report the case. How to identify auto insurance fraud Behavior has always been a difficult problem for insurance practitioners and the field of insurtech. [0003] Traditional anti-fraud technologies mostly rely on rule engines and statistical models. During the deployment and implementation of rule engine...

AI Technical Summary

Problems solved by technology

[0003] Traditional anti-fraud technologies mostly rely on rule engines and statistical models. During the deployment and implementation of rule engines, business experts need to constantly update the rules, which consumes a lot of manpower. Statistical models are difficult to deeply explore the relationship between model factors, especially It is a high-level relationship, so its processing capacity is greatly limited. At the same time, there are regional differences and time differences in insurance fraud. As a result, there is currently no more complete method that can accurately and efficiently detect insurance claim fraud.

Images