Skip to content

Latest commit

 

History

History
14 lines (9 loc) · 2.83 KB

README.md

File metadata and controls

14 lines (9 loc) · 2.83 KB
Raw

Webinars on Cognitive AI

This page is intended to support arranging webinars on behalf of the Cognitive AI Community Group. These webinars would be open to all, and not just the members of the Cognitive AI Community Group. The general aim is to draw attention to the potential of Cognitive AI, to present existing work, and to encourage discussion on research directions, software and educational resources, as well as dissemination and cross-coupling with other communities. Webinars should be arranged at times convenient to as many people as possible, and will be recorded for public access.

Initial Webinar

Please feel free to provide pull requests for updating this section with your suggestions for the initial webinar.

  • A presentation by Dave Raggett on his work on plausible reasoning, including the web-based demo for plausible reasoning (inspired by Allan Collins), analogical reasoning (inspired by Dendre Gentner), fuzzy reasoning (inspired by Lotfi Zadeh), and the potential for fuzzy quantifiers and modifiers as an extension to traditional logic\
  • A presentation by Paola Di Maio (AI KR CG) on the intersection and co-evolution of Cognitive AI, Knowledge Representation and Machine Learning, expounding the system level knowledge representation and the need for KR to ensure reliablty transparency and accountability of CogAI. The presentation discusses the role of KR in AI, the risks arising from lack of KR, as well as the future of KR in ML
  • A presentation by Pierre Monnin of (i) how domain knowledge can improve the performance of a matching approach based on Graph Convolutional Networks in Knowledge Graphs (Monnin et al., Semantic Web Journal, 2022), and (ii) how domain knowledge offers additional evaluation perspectives for Machine Learning models with the Sem@K metric (Hubert et al., EKAW 2022, Hubert et al., DL4KG@ISWC 2022)
  • An open discussion moderated by Dave Raggett on recent work on deep learning, e.g. OpenAI's DALL•E 2, Hugging Face's Stable Diffusion, Google's DreamBooth, Hugging Face's BLOOM, current limitations, and the challenges for achieving scalable learning for machine-based reasoning, see e.g., fundamental limitations/drawbacks of deep learning everyone should know (Sharad Joshi on medium.com)
  • others to be added