A Method for Identifying Malicious Activity in Collaborative Systems with Maps

In this paper we describe the method we have created for the purpose of identifying misbehavior of users who intends to generate false trends in digital maps. Basically, the idea is to identify patterns of communities of users who strongly contribute with reports that lead a particular geographic area to be considered a hot spot. The association between hot spots, computed from Kernel Density Estimation techniques, and the methods for identifying communities in social networks is the main innovation of the method proposed. A multi-agent system was built in order to simulate several scenarios of malicious activities. This method has shown to be effective for alerting the possibility of malicious activity in a real system.