More than 900 individuals from the international academic community, the industrial world, and the general public attended the virtual workshop on neural-symbolic integration entitled When deep learning meets logic. The workshop was co-organised by Loizos Michael (computational cognition lab, Open University of Cyprus), Efi Tsamoura (Samsung AI Centre, Cambridge), Vaishak Belle (University of Edinburgh), and Phokion Kolaitis (University of California, Santa Cruz and IBM Research) and was sponsored by Samsung Research.
The recent wide-applicability of deep learning techniques in dealing with unstructured data like text, images, and videos at a massive scale has intensified the efforts to integrate this line of Artificial Intelligence (AI) research with another prominent AI paradigm: the use of symbolic knowledge and logic reasoning. Although these two strands of AI research have for the most part evolved independently, their interplay has shown a potential to give rise to a new computational paradigm in which symbolic knowledge extends and enhances the capabilities of deep learning systems, while also offering a path towards the grounding of symbols and the induction of knowledge from sensory data.
The synergy of these two paradigms was the timely topic of the workshop When deep learning meets logic. Featuring prominent speakers from world-class academic and industrial institutes, the workshop focused on exploring state-of-the-art solutions and their maturity for adoption by the industrial sector, and also on identifying some of the big questions that are still open for debate and research.