Graph Mining

Frequent Subgraph Mining in Dynamic Graphs

Contact Student: Abhik Ray, abhik.ray@email.wsu.edu

Details:

Previous efforts in graph mining have been largely devoted to finding frequent subgraphs in static graphs from various domains such as social networks, biological networks, chemical compounds etc. Most real world graphs and networks are however dynamic in nature with nodes, edges and labels being added, deleted and modified over time. My project therefore tries to solve the problem of finding frequent subgraphs in such evolving graphs. To make it sound cool I work with Facebook data.

Datasets:

- https://socialnetworks.mpi-sws.org/

- SNAP at Stanford: https://snap.stanford.edu/

- U Florida Sparse Matrix Collection: https://www.cise.ufl.edu/research/sparse/matrices/index.html

Publications:

- Dynamic Tensor Analysis: www.cs.cmu.edu/~christos/PUBLICATIONS/kdd06DTA.pdf

- Dynamics of large networks by J. Leskovec. PhD Dissertation, Machine Learning Department, Carnegie Mellon University, Technical report CMU-ML-08-111, 2008.

Other Links

- NetLogo Software: https://ccl.northwestern.edu/netlogo/

- NetworkX: https://networkx.lanl.gov/

- Academia.edu homepage: https://wsu.academia.edu/AbhikRay