HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation
Conference papers

Learning Abstracted Non-deterministic Finite State Machines

Abstract : Active automata learning gains increasing interest since it gives an insight into the behavior of a black-box system. A crucial drawback of the frequently used learning algorithms based on Angluin’s $$L^*$$L∗ is that they become impractical if systems with a large input/output alphabet are learned. Previous work suggested to circumvent this problem by abstracting the input alphabet and the observed outputs. However, abstraction could introduce non-deterministic behavior. Already existing active automata learning algorithms for observable non-deterministic systems learn larger models if outputs are only observable after certain input/output sequences. In this paper, we introduce an abstraction scheme that merges akin states. Hence, we learn a more generic behavioral model of a black-box system. Furthermore, we evaluate our algorithm in a practical case study. In this case study, we learn the behavior of five different Message Queuing Telemetry Transport (mqtt) brokers interacting with multiple clients.
Complete list of metadata

Contributor : Hal Ifip Connect in order to contact the contributor
Submitted on : Thursday, May 27, 2021 - 4:42:58 PM
Last modification on : Thursday, May 27, 2021 - 4:58:37 PM
Long-term archiving on: : Saturday, August 28, 2021 - 8:00:02 PM


 Restricted access
To satisfy the distribution rights of the publisher, the document is embargoed until : 2023-01-01

Please log in to resquest access to the document


Distributed under a Creative Commons Attribution 4.0 International License



Andrea Pferscher, Bernhard Aichernig. Learning Abstracted Non-deterministic Finite State Machines. 32th IFIP International Conference on Testing Software and Systems (ICTSS), Dec 2020, Naples, Italy. pp.52-69, ⟨10.1007/978-3-030-64881-7_4⟩. ⟨hal-03239824⟩



Record views