Learning Node Label Controlled Graph Grammars (Extended Abstract)

Within the data mining community there has been a lot of interest in mining and learning from graphs (see [1] for a recent overview). Most work in this area has has focussed on finding algorithms that help solve real-world problems. Although useful and interesting results have been obtained, more fundamental issues like learnability properties have hardly been adressed yet. This kind of work also tends not to be grounded in graph grammar theory, even though some approaches aim at inducing grammars from collections of graphs. This paper is intended as a step towards an approach that is more theoretically sound. We present results concerning learnable classes of graph grammars.