University of Luxembourg, together with University of Turin and Stanford University, is carrying out research projects in Natural Language Processing applied to legal documents with Nuance Communication, specifically with dr Valeria De Paiva, working at the Artificial Intelligence and Natural Language Understanding Sunnyvale Lab.

Currently, the research activity focuses on devising a new deep learning model for textual entailment in the cybersecurity domain. The project aims at deriving whether NIST controls for security and privacy for federal information systems and organizations entail ISO/IEC 27000 controls and viceversa.

Contrary to standard deep learning approaches for textual entailment, which directly act on text, the present project will try to investigate deep learning techniques on flat semantic representations, such as those built in reified Input/Output logic.