The development of complex and dependable systems like autonomous vehicles relies increasingly on the use of systems modeling language (SysML). In fact, SysML has become a de facto standard for systems engineering. With model-driven engineering, a SysML model serves as a reference for the early defect detection of the system under design: the earlier the errors are detected, the less is the cost of handling the errors. Mutation testing is a fault-based technique that has recently seen its applications to SysML behavioral models (e.g., state machine diagrams). Specifically, a system's state-transition design can be fed to a model checker where mutants are automatically generated and then killed against the desired design specifications (e.g., safety properties). In this paper, we present a novel approach based on process mining to improve the effectiveness and efficiency of the SysML mutation testing based on model checking. In our approach, the mutation operators are applied directly to the state machine diagram. These mutants are then fed as traces into a process mining tool and checked according to the event logs. Our initial results indicates that the process mining approach kills more mutants faster than the model checking method.
Aurek Chattopadhyay, Reagan Maddox, Glen Horton, Nan Niu, Ganesh Malla, Tanmay Bhowmik, Jianzhang Zhang, and Juha Savolainen, Completeness of Natural Language Requirements: A Comparative Study of User Stories and Feature Descriptions (submitted to REFSQ 2023: https://2023.refsq.org)
Artifacts of the paper entitled:
Prompting Creative Requirements via Traceable and Adversarial Examples in Deep Learning
Authors: Hemanth Gudaparthi, Nan Niu, Boyang Wang, Tanmay Bhowmik, Hui Liu, Jianzhang Zhang, Juha Savolainen, Glen Horton, Sean Crowe, Thomas Scherz and Lisa Haitz
To appear in the Proceedings of the 31st IEEE International Requirements Engineering Conference (RE 2023 https://conf.researchr.org/home/RE-2023)