Fraud detection and management strategies in the insurance organizations using data mining techniques

Today, fraud, which dates back to the size of human life, is a multi-million dollar business worldwide, and its financial size is increasing day by day. In recent years, the development of new technologies has opened many ways for fraudsters and criminals to commit fraud. Creating a new information system, in addition to all its benefits and benefits, may offer more opportunities for offenders to commit fraud.  Identification of the critical points of the occurrence of fraud in the first step. Identifying the fractious sectors of the Social Security Organization by completing the questionnaire, interviewing and reviewing cases, the prioritization and related data were extracted from the relevant database. Primary refinement data and key variables were identified. All possible scenarios from the two existing "cheating" and "healthy" situations are designed to prepare the test and test data sheets in the form of Excel files and to identify the patterns of behavior of the insured organizations of the Social Security Organization in order to exploit the resources available to the data mining algorithms. The results of the implementation of artificial neural network algorithms, decision trees and the nearest designer neighbor were identified in the form of three separate tests by simulation software and the best algorithm for implementing the final SQL code. In the final section of the research, the manufacturer's suggestions Reforming and preventing fraud in the social security organization It is described in detail.