We have implemented the state-enhanced k-tail (SEKT) and the trace-enhanced MTS inference (TEMI) algorithms in Java. Both Java binary and source code of the implementation are available below.
Java Binary Usage:
Use the following command in the directory containing the file MTSInference.jar:
java -jar MTSInference.jar tracefile invariantfile packagename classname outputpath
The following are ground-truth models of the nine libraries we used to evaluate k-tail, CONTRACTOR++, SEKT and TEMI algorithms. Ground-truth models represent the exception-free behavior of the libraries. We took these models from related work ,  and modified them to correct imprecisions. All model files are in the FSP format used by the LTSA and MTSA model checking tools.
 V. Dallmeier et al. Automatically generating test cases for specification mining. IEEE TSE, 38(2), 2011.
 M. Pradel et al. A framework for the evaluation of specification miners based on finite state machines. In Proc. of ICSM, 2010.
The traces used to evaluate the inference algorithms can be found at testing traces. The traces are generated by Daikon and are read as the first step of the above implementation.