Improving Protocol Vulnerability Discovery via Semantic Interpretation of Textual Specifications [59F-A52]
Samuel Jero, Maria Leonor Pacheco, Dan Goldwasser, Cristina Nita-Rotaru
This work seeks to automatically extract the information about network protocols that is present in their natural language specification documents and to learn structured information about the protocol. In particular, we focus on extracting message formats, protocol state machines, constraints, and invariants. We then apply this information to the automated testing and model checking of network protocols, with the goal of providing a completely automated testing solution for protocol implementations.