print · source · login   



Tomte-0.4 can learn register automata with fresh output values which it handles using a Determinizer function. Check the Tomte-0.4 Fresh ICTAC publication for more details on our algorithm. In a learning example for the MultiLogin Bennchmark Tomte 0.4 was used to infer models for instantiations of Multi-Login Systems with 1, 2 and 3 maximum registered users. Improvements in Tomte 0.4 also give better performance in learning register automata. You can compare the old numbers in Tomte-0.3 ISOLA14 paper with the improved numbers in Tomte-0.4 Fresh ICTAC publication. Tomte 0.41 can learn the same type of systems but the implementation of the algorithms got improved and is used for the Tomte-0.41 Fresh Full publication where we compare Tomte 0.41 with RALib which is a new learning algorithm specially for register automata.
Also see: Tomte-0.41 Documentation


Tomte-0.3 can learn register automata. We did a comparison of Tomte-0.3 and Learnlib for learning register automata for result data see Tomte-0.3 ISOLA14 paper.
Also see: Tomte-0.3 Documentation


Tomte-0.2 can learn SUTs that may only remember the last and first occurrence of a parameter. Tomte-0.2 is an improved implementation of Tomte-0.1 in which we added the SUT tool. We used Tomte-0.2 in the paper "Learning and Testing the Bounded Retransmission Protocol" and in the the improved journal version of the paper "Improving Active Mealy Machine Learning for Protocol Conformance Testing" .
Also see: Tomte-0.2 Documentation


Tomte-0.1 can learn SUTs that may only remember the last and first occurrence of a parameter. We are not able yet to learn timed systems. Also see the paper "Automata Learning Through Counterexample-Guided Abstraction Refinement" [CEGAR12].
Also see: Tomte-0.1 Documentation.