ITALIA
Integrating Testing And
Learning of
Interface Automata
STW project 11763
Research Summary
The ITALIA project deals with the design of algorithms that will allow computers to learn complex state diagrams by providing inputs and observing outputs. The state diagrams that can be learned by current techniques have at most 30.000 states. In contrast, the state diagrams that govern the behavior of computing based systems (defined using dozens of state variables) typically have more than 101000 states. The research objective of ITALIA is to construct a tool set that will allow us to learn - routinely and fully automatically - state diagrams with up to 40 state variables. The ITALIA project will focus on the utilisation of model inference technology within the area of testing: once we have learned a model of a software component, we will use model checking technology to analyze this model (e.g. to detect security vulnerabilities) and the technology of model based testing to automatically infer test suites. Using these test suites we can then check, for instance, (a) whether no new faults have been introduced in a modified version of the component (regression testing), (b) whether an alternative implementation by some other vendor agrees with a reference implementation, or (c) whether some new implementation of legacy software is correct.Project Leader
Prof.dr. F.W. VaandragerInstitute for Computing and Information Sciences
Mailbox 47, Faculty of Science
Radboud University Nijmegen
Heijendaalseweg 135, Huygens Building, room 02.622
6525 AJ Nijmegen
The Netherlands
Project Status
Starting date: 01-01-2012Ending date: 31-12-2015
Keywords
software, state machines, machine learning, model inference, testing, embedded controllers, network protocolsZijbalk
News
- Invited talk of Frits Vaandrager at TNO-ESI symposium, April 19, 2016, Eindhoven.
- Paper on TCP accepted at CAV'16!!!
- Two papers on model learning accepted at iFM
- Invited talk of Frits Vaandrager at second Dutch national symposium on software engineering (SEN), January 21st 2016, Amsterdam.
- New paper on use of log-based metrics to enhance automata learning!
- New paper on industrial case study at Philips!
- Our group gave two presentations at the Nederlandse Testdag, October 6, 2015, Eindhoven.
- Our latest paper on Tomte 0.4 has been presented at ICTAC'15 in Colombia.
- Our work on applying automata learning to embedded control software has been presented at ICFEM'15 in Paris.
- On January 10, 2015, Frits Vaandrager appeared on Dutch National TV (EenVandaag) to tell about the scientific importance of Alan Turing, in connection with the movie "The Imitation Game".
- On January 7, 2014, the 5th meeting of the ITALIA user committee took place in Nijmegen.
- On October 27, 2014, Fides Aarts successfully defended her PhD thesis Tomte: Bridging the Gap between Active Learning and Real-World Systems
- Two ITALIA papers were presented at the ISOLA'14
- Tomte 0.3 released with its own new website
- On September 12, 2014, Paul Fiterau and Frits Vaandrager presented a paper at FMICS'14 in which they show that both Windows and Linux violate the TCP standard.
- Frits Vaandrager will present several lectures on automata learning at ESSCaSS 2013, August 18-22, Estonia.
- Sicco Verwer received a Veni grant from STW.
- Frits Vaandrager received a grant from NWO for research on automata learning.
- Sicco Verwer and Frits Vaandrager received a grant from NWO to learn extended state machines for malware analysus.
- Frits Vaandrager has presented a keynote lecture on automata learning at ICTSS'12, November 19-21, 2012, Aalborg, Denmark
- The second ITALIA user committee meeting was held on November 15, at Chess eT International B.V., in Haarlem
- Fides
Aarts receives Frye Grant
- On September 28, at 11.00am Wouter Smeenk will present his MSc thesis Applying Automata Learning to Complex Industrial Software in HG00.065. This work was carried out at Océ Research within the context of the ITALIA project.
- Tomte 0.2 released
- New Sut tool released
To simulate an SUT from a state machine model.