Homeland security requires that we draw inferences about activities from massive flows of data arriving continuously over time. Buried in these data are patterns and behaviors that are changing, often quite quickly. DyDAn researchers will develop novel technologies to find patterns and relationships in dynamic, nonstationary, and sometimes massive data sets. Our work involves Information Management and Knowledge Discovery as well as topics in the foundations of discrete mathematics, and it ranges over a wide variety of applications from early warning of disease outbreaks to responses to natural disasters.
Fliers describing a sample of ongoing projects in homeland security can be found here.
DyDAn's research spans two primary areas of theoretical research: (1) Analysis of Large, Dynamic Multigraphs and (2) Continuous, Distributed Monitoring of Dynamic, Heterogeneous Data. We plan several research projects under these two themes, as well as several more applied projects:Projects Involving Analysis of Large, Dynamic Multigraphs:
- Analysis of Large, Dynamic Multigraphs Arising from Blogs
- Universal Information Graphs
- Statistical and Graph-Theoretical Approaches to Time-Varying Multigraphs
- Adding Semantics to and Interconnecting Semantic Graphs
- Algorithms for Identifying Hidden Social Structures in Virtual Communities
- Continuous Distributed Data Stream Monitoring
- Message Filtering and Entity Resolution/Author Identification (machine learning)
- Dynamic Similarity Search in Multi-Modal Data
- Optimization and Learning
- Privacy-Preserving Data Analysis
- Potential Uses of Entropy in Biosurveillance
- Enabling and Enhancing Crime Prevention and Analysis at the Port Authority of New York and New Jersey
- Sensor Management for Nuclear Detection