RWTH Aachen University
In this paper, the authors study a graph kernel for RDF (Resource Description Framework) based on constructing a tree for each instance and counting the number of paths in that tree. In their experiments this kernel shows comparable classification performance to the previously introduced intersection subtree kernel, but is significantly faster in terms of computation time. Prediction performance is worse than the state-of-the-art Weisfeiler Lehman RDF kernel, but their kernel is a factor 10 faster to compute. Thus, they consider this kernel a very suitable baseline for learning from RDF data.