||June 9, 2006
||Dr. Tamas Horvath, Department of Computer Science III, University of Bonn
||Mining and Learning from Graph Structured Data
|| In recent years, there has been an increasing interest in mining and
from graph structured data.
This field of research is motivated by various application areas, such as
analysis of chemical graphs in pharmaceutical applications, graph structures
of the World Wide Web, or social networks.
Depending on the specification of the instance space (e.g., set of disjoint
set of m-tuples of vertices of a graph etc.), the matching operator
(e.g., subgraph isomorphism, homomorphism etc.), and the computational task
(e.g., learning predictive models, enumerating local patterns etc.) various
settings can be defined for mining and learning from graph structured data.
In this talk, we present theoretical and/or empirical results with
for several such problem settings.