Model-based fraud detection in growing networks

People share opinions, exchange information, and trade services on large, interconnected platforms. As with many new technologies these platforms bring with them new vulnerabilities, often becoming targets for fraudsters who try to deceive randomly selected users. To monitor such behavior, the proposed algorithm evaluates structural anomalies that result from local interactions between users. In particular, the algorithm evaluates the degree of membership to well-defined communities of users and the formation of close-knit groups in their neighborhoods. It identifies a set of suspects using a first order approximation of the evolution of the eigenpairs associated to the continuously growing network. Within the set of suspects, the algorithm them locates fraudsters based on deviations from the expected local clustering coefficients. Simulations illustrate how incorporating asymptotic behavior of the structural properties into the design of the algorithm allows us to differentiate between the aggregate dynamics of fraudsters and regular users.