作者: Gregor Gössler , Daniel Le Métayer
DOI: 10.1016/J.SCICO.2015.06.010
关键词: Blame 、 Safety property 、 TRACE (psycholinguistics) 、 Component (UML) 、 Counterfactual thinking 、 Computer science 、 Computer security 、 Outcome (game theory) 、 Liability 、 Causality
摘要: In component-based safety-critical embedded systems it is crucial to determine the cause(s) of violation a safety property, be issue precise alert, steer system into safe state, or liability component providers. this paper we present an approach blame components based on single execution trace violating property P . The diagnosis relies counterfactual reasoning ("what would have been outcome if C had behaved correctly?") distinguish failures that actually contributed from little no impact