kLog: A Language for Logical and Relational Learning with Kernels
Preprint
Abstract
kLog is a logical and relational language for kernel-based learning. It allows users to specify logical and relational learning problems at a high level in a declarative way. It builds on simple but powerful concepts: learning from interpretations, entity/relationship data modeling, logic programming and deductive databases (Prolog and Datalog), and graph kernels. kLog is a statistical relational learning system but unlike other statistical relational learning models, it does not represent a probability distribution directly. It is rather a kernel-based approach to learning that employs features derived from a grounded entity/relationship diagram. These features are derived using a novel technique called graphicalization: first, relational representations are transformed into graph based representations; subsequently, graph kernels are employed for defining feature spaces. kLog can use numerical and symbolic data, background knowledge in the form of Prolog or Datalog programs (as in inductive logic programming systems) and several statistical procedures can be used to fit the model parameters. The kLog framework can -- in principle -- be applied to tackle the same range of tasks that has made statistical relational learning so popular, including classification, regression, multitask learning, and collective classification.Keywords
All Related Versions
This publication has 25 references indexed in Scilit:
- Practical graph isomorphism, IIJournal of Symbolic Computation, 2014
- Fast learning of relational kernelsMachine Learning, 2010
- Structured machine learning: the next ten yearsMachine Learning, 2008
- A parametric filtering algorithm for the graph isomorphism problemConstraints, 2008
- Convex multi-task feature learningMachine Learning, 2008
- Graph kernels for chemical informaticsNeural Networks, 2005
- Solving the multiple instance problem with axis-parallel rectanglesArtificial Intelligence, 1997
- Inverse entailment and progolNew Generation Computing, 1995
- Applications of stochastic context-free grammars using the Inside-Outside algorithmComputer Speech & Language, 1991
- Isomorphism of graphs of bounded valence can be tested in polynomial timeJournal of Computer and System Sciences, 1982