Securboration, Inc.

Bruce R McQueary

Craig T Hagan

Lee Krause

James B Schneider

Brian Warhover

Anthony Stirtzinger

Keith Shapiro

Carlos E Guestrin

Sean K Thornton

Peter M LaMonica

Craig S Anken

Robert G Asfar

Methods and systems for classifying social media users. The system computes a plurality of subgraphs from a user's social graph network and considers which types of subgraphs are overly represented in the user's social network to determine whether a user belongs to a certain class. The system may also consider features based on metadata of the user's network and social interactions occurring in the user's network.

Classifying social media users

Systems, methods, and apparatus for identifying anomalous behavior are provided. For example, a method may include receiving raw data, generating a behavior profile for the entity based on the raw data, receiving comparison data, determining whether the comparison data deviates from a pattern of behavior defined in the behavior profile, and identifying the comparison data as anomalous behavior when the comparison data deviates from the pattern of behavior. In one embodiment, the raw data includes recorded activity for the entity. In one embodiment, the behavior profile defines a pattern of behavior for the entity. In one embodiment, a countermeasure is performed upon identifying anomalous behavior. The countermeasure may include at least one of revoking the entity's credentials, denying the entity access to a resource, shutting down access to a port, and denying access to the entity. The method may further include providing a report of the anomalous behavior.

Cyber-semantic account management system

Methods and systems for motif recognition in graph sets. Distribution scores comparing the canonical representations of a compressed graph in the set of graphs versus the canonical representations of a compressed graph in a reference graph are used to determine the significance of the compressed graph.

Motif recognition

Methods and systems for node and link detection in social network analysis. Interactive noise reduction allows reduction of the data set under analysis to enable substantially real time detection of links and nodes.

Methods and systems for node and link identification

Growing community of inventors

Indialantic, FL, United States of America

Bruce R McQueary