Learning Abstracted Non-deterministic Finite State Machines

Andrea Pferscher*, Bernhard Aichernig

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-review


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 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.

Original languageEnglish
Title of host publicationTesting Software and Systems - 32nd IFIP WG 6.1 International Conference, ICTSS 2020, Proceedings
Subtitle of host publication32nd IFIP WG 6.1 International Conference, ICTSS 2020, Naples, Italy, December 9-11, 2020, Proceedings
EditorsValentina Casola, Alessandra De Benedictis, Massimiliano Rak
Number of pages18
ISBN (Print)978-3-030-64880-0
Publication statusPublished - Dec 2020
Event32nd IFIP International Conference on Testing Software and Systems: ICTSS 2020 - Virtuell, Italy
Duration: 9 Dec 202011 Dec 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12543 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference32nd IFIP International Conference on Testing Software and Systems
Abbreviated titleIFIP-ICTSS 2020


  • Active automata learning
  • Model inference
  • Non-deterministic finite state machines
  • MQTT

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fields of Expertise

  • Information, Communication & Computing


Dive into the research topics of 'Learning Abstracted Non-deterministic Finite State Machines'. Together they form a unique fingerprint.

Cite this