My research is in learning and discovering meaningful network representations that describe complex, time-evolving systems. For example, a classic neuroimaging task is to convert time-varying fMRI scans of a subject’s brain into a network representation. How to do this is not immediately clear. Because all eventual analyses and conclusions depend on the network representation of the brain, finding the “right” representation is crucial.
There are many other domains in which sequential or time-evolving data may be represented as a graph, enabling network-specific insights that individual observations alone may not give. Past and current projects explore professional career trajectories, personalized web search, email communications, and more. Overall, my work is most related to network inference and discovery, data summarization, and sequence learning – falling somewhere between data mining, machine learning, and network science.
For a full list, see my CV.
Career Transitions and Trajectories: A Case Study in Computing
Tara Safavi, Maryam Davoodi, Danai Koutra
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)
Additional presentation at the KDD’18 BigScholar Workshop
Graph Summarization Methods and Applications: A Survey
Yike Liu, Tara Safavi, Abhilash Dighe, Danai Koutra
ACM Computing Surveys (CSUR)
Scalable Hashing-Based Network Discovery
Tara Safavi, Chandra Sripada, Danai Koutra
IEEE International Conference on Data Mining (ICDM)
Full paper – 9% acceptance rate
Nominated for best paper
Reducing Million-Node Graphs to a Few Structural Patterns: A Unified Approach
Yike Liu, Tara Safavi, Neil Shah, Danai Koutra
ACM SIGKDD Workshop on Mining and Learning With Graphs (KDD MLG)