Sigma Identity Fraud machine learning models are purpose-built for specific industries, trained with feedback data from a consortium of clients to tackle targeted fraud patterns and produce real-time actionable risk scores and reason codes.
Socure ID+ email risk models predict the risk of an email and its correlation to an identity considering hundreds of data elements. Predictive factors include deliverability, age of the email, domain type, and alternative emails linked to an identity.
Socure ID+ phone risk models predict the risk of a phone number utilizing various risk intelligence and phone number attributes. Predictive factors include how often the phone number has been ported, length of subscriber tenure, and if the number is a VoIP.
Socure ID+ address risk models predict the risk of an address utilizing various risk intelligence and residency attributes. Predictors factors include deliverability, length of time at an address, property type, as well as links between an identity and alternative addresses.
Socure’s consortium database of known first- and third-party perpetrators and fraud outcomes, tagged by industry vertical or segment, is an additional source of data in the decisioning process to accept, reject, or review consumer applicants.