This project is for research for predicting and detecting social network membership based on automated forensic analysis, unsupervised learning techniques, and statistical inference. The research focuses on establishing network membership from hard drive data, using a corpus of 750 seized hard drives already obtained by the PI and senior researcher. The project will conduct novel research to detect social networks implicit in common information links between seized hard drives and related media. The work will broadly apply to any large data collection in multiple domains.