 |
Center for Nonlinear Studies |
Many problems can be modeled with graphs; entities are represented as vertices and relationships between entities are represented as edges. In graph data mining, we seek to find interesting patterns in a single graph or a collection of graphs. These patterns can be structural and/or semantic. Graph data mining is a specialized area of data analytics, wherein one analyzes a set of raw data and makes some conclusions. However, there are different forms of data analytics. In diagnostic analytics, we want to understand why something happened. In predictive analytics, the objective is to understand what is likely to happen in the future. Prescriptive analytics is subtly different; the goal is to discover what we can do to make something happen, or to prevent something from happening. In this talk, we will look at three graph data mining applications that range from software engineering to cyber-physical systems. Each of these applications employs diagnostic, predictive, and/or descriptive analytics.
Dr. Jennifer Leopold is currently an Associate Professor of Computer Science at Missouri University of Science and Technology in Rolla, Missouri. She received her B.S. in Mathematics, M.S. in Computer Science, and Ph.D. in Computer Science, all from the University of Kansas. Her research interests range broadly in end-user programming environments, with particular focus on data accessibility and analysis, and scientific visualization. Early in her career, she pursued those interests through research projects in the field of bioinformatics, particularly with respect to biodiversity and morphology. Her current research projects focus on data analytics, applying graph data mining to problems in software engineering, communication networks, and cyber-physical systems.
Host: Beth Hornbein