On Mining Graph Patterns

Prof. Philip S. Yu

Professor and Wexler Chair in Information Technology, Department of Computer Science University of Illinois at Chicago

With ever-increasing amounts of graph data from disparate sources, there has been a strong need for exploiting significant graph patterns with user-specified objective functions to define the feature space for graph classification in a large graph database. Most objective functions are not anti-monotonic, which could fail all of frequency-centric graph mining algorithms. In this talk, we examine the issue on general mining method aiming to discover the most significant patterns directly. However, in many applications, the labels of graph data are very expensive or difficult to obtain, while there are often copious amounts of unlabeled graph data available. We will also study the problem of semi-supervised feature selection for graph classification.