Abstract: In some aspects, a computing system is configured to use graph-based techniques to detect synthetic identities. The computing system can generate a collection of graphs based on account data and transaction data for online entities. The collection of graphs includes multiple graph communities, each graph community including nodes and edges. Each node represents a user and an edge between a first node and a second node indicates a user represented by the second node is an authorized user of the user represented by the first node. The computing system can identify a clique graph community in the collection of graphs and compare the identified clique graph community with a known clique graph community that includes synthetic identities. The computing system can determine nodes in the identified clique graph community to be synthetic identities based on determining that the identified clique graph community is equivalent to the known clique graph community.

Abstract: Certain aspects involve using a set of machine learning modeling models for predicting attempts to tamper with records using a fraudulent dispute. A tampering prediction system receives a request from a target entity to modify event data for a historical event, including information about the target entity and the event. The system generates a first score by applying a first set of machine learning models to the information from the request and information about the target entity obtained from a database. They system computes a second score by applying a second machine learning model to event data retrieved from the database. The second machine learning model has been trained using labeled training data and is augmented with a model that has been trained using unlabeled training data. The system generates an overall score for the request based on the first score and the second score.