This book describes exciting new opportunities for utilizing robust graph representations of data with common machine learning algorithms. Through the use of graph distance - a relatively new approach for determining graph similarity - the authors show how the well-known algorithms, like k-means and k-nearest neighbors classification, can be easily expanded to work with vector graphics instead. Several methods of representing web document content by the graph introduced; interesting feature of this representation is that they allow for the calculation of the distance polynomial time, something that usually is NP-complete problems when using the graph. The experimental results are reported for both clustering and classification in the three web document collections, using a variety of graphical representations, measure distances, and algorithm parameters.