نوع مقاله : مقاله علمی

نویسندگان

• اسماء حمزه
• محمدجواد نجفی آرانی

کلیدواژه‌ها

• بیمه خودرو
• نظریه گراف
• توزیع پواسون
• برچسب‌گذاری

موضوعات

• ریاضیات مالی/ کاربردی

عنوان مقاله [English]

A mathematical model for identifying and validating suspicious clusters associated with organized fraud in auto insurance

نویسندگان [English]

• Asma Hamzeh
• Mohammad Javad Nadjafi-Arani

چکیده [English]

BACKGROUND AND OBJECTIVES: Insurance fraud is a common challenge in the industry, leading to significant losses both in terms of financial interests and public trust. Financial and monetary institutions are keenly seeking to accurately identify the activities of fraudsters and fraudsters. Due to its direct effect on serving the clients of institutions, this will lead to the reduction of operating costs, gaining the trust of other insurers, and maintaining and improving the market share of insurers as reliable financial service providers. One of the most prevalent forms of fraud occurs in auto insurance, where organized and opportunistic fraudulent activities are widespread. Intentional accidents, especially those involving groups, staged injuries, and orchestrated scenes, are among the common fraudulent practices in this domain.
METHODS: One of the techniques used to detect fraud is network analysis. In the network analysis, the communication between people and different real and legal personalities are evaluated and new dimensions of these communication are identified. The objective of this paper is to introduce a mathematical model based on graph theory for identifying suspicious clusters associated with organized fraud. In our research, we first introduce a network called the “accident network” using graph theory. We demonstrate that this network exhibits characteristics of a random graph. Suspicious clusters within this network are then identified using an algorithm based on graph theory. Subsequently, we examine the occurrence probability of such clusters in a random accident network by defining a binomial distribution over its edges.
FINDINGS: This process leads to assigning a label (indicating fraudulent or non-fraudulent) to each accident and individual. Considering the structure of the algorithm and its complexity, we can conclude that the proposed algorithm is simply capable of analyzing a lot of data.
CONCLUSION: Investigating this topic enables insurers to tailor different policies based on the labels assigned to individuals or accidents, ultimately aiming to reduce financial losses and enhance public trust.

کلیدواژه‌ها [English]

• Car insurance
• Graph theory
• Poisson distribution
• labeling