Learning and Model-checking Networks of I/O Automata

We introduce a new statistical relational learning (SRL) approach in which models forstructured data, especially network data, are constructed as networks of communicatingnite probabilistic automata. Leveraging existing automata learning methods from the areaof grammatical inference, we can learn generic models for network entities in the form ofautomata templates. As is characteristic for SRL techniques, the abstraction level aordedby learning generic templates enables one to apply the learned model to new domains. Amain benet of learning models based on nite automata lies in the fact that one can analysethe resulting models using formal model-checking techniques, which adds a dimension ofmodel analysis not usually available for traditional SRL modeling frameworks