Visualization of large relational data sets and clusters as graphs and maps

Yifan Hu
Seminar

In the first part of this talk, we will discuss challenges in dealing
with large graphs in general, including the task of visualizing all of
the sparse matrices in the University Florida Sparse Matrix
Collection. While traditional graph visualization methods can be
invaluable in getting an overall sense out of large data sets, they
are not as helpful in conveying the underlying structural information,
clusters, and neighborhoods. In the second part of this talk, we
describe an algorithm, GMap, for visualizing graphs as maps. GMap
overcomes some of the shortcomings with the help of the geographic map
metaphor. The effectiveness this algorithm is illustrated with
visualization examples from several domains, namely Netflix movies, TV
shows, Amazon books, and last.fm music. Some results of this talk can
be found at http://www.research.att.com/~yifanhu/gallery.html