TITLE: Tools for non-traditional data mining

SPEAKER: Christos Faloutsos (CMU)

ABSTRACT: How can we find patterns on multimedia, or multi-modal data? for example, we have video, with still frames, audio, text, and motion; or a collection of images with labels, and we want to automatically label a new image. Another related area is sensor mining: given a large collection of sensors, producing numerical measurements, how can we find correlations between them, in a streaming, 'any-time' fashion?
We present recent attempts to these problems. The idea in the first is to turn the problem into a graph problem, and eventually apply existing graph algorithms (like PageRank), or develop new ones. For the second problem, we provide an array of methods that can spot patterns in a single sequence, as well as across multiple sequences.

BIO: Christos Faloutsos is a Professor at Carnegie Mellon University. He has received the Presidential Young Investigator Award by the National Science Foundation (1989), five ``best paper'' awards, and several teaching awards. He is a member of the executive committee of SIGKDD; he has published over 130 refereed articles, one monograph, and holds five patents. His research interests include data mining for streams and networks, fractals, indexing methods for spatial and multimedia bases, and data base performance.