This tool will be/is included in LearnLib v0.17

---

This tool introduces the concept of Systems of Behavioral Automata (SBAs), an extension to SPAs that support prefix-closure. For running the tool, you need

• a working JDK (8+) installation
• a working Maven installation

---

The tool consists of two modules -- learner and benchmark.

• learner

  The learner module contains the core components of the learning algorithm.

  • The SBALearner class resembles the main class of the learning algorithm and integrates into the LearningAlgorithm framework of the LearnLib.
  • The StackSBA class resembles a stack-based implementation that accepts the (instrumented) language of an SBA.
  • The config package contains several predefined adapters that can be used to configure, which learning algorithms the SBALearner should use for learning the individual sub-procedures of the system under learning.

• benchmark

  The benchmark module contains a Main class that resembles the entry point to the benchmarks. You can either start the benchmark from an IDE or build an executable benchmark JAR via the following steps:

  • Run mvn clean package,
  • In the benchmark/target/benchmark/directory you will find a learnlib-sba-benchmark-1.0-SNAPSHOT-jar-with-dependencies.jar which can be executed with java -jar path/to/jar
    • Once started, the benchmark will create two files (sba.csv, output.log) in the directory from which you started the benchmark.
    • The benchmarks run in parallel. Depending on how many cores your system has, the process may require multiple GBs of RAM.