September 08, 2021
ML2R scientist Dr. Tiansi Dong is the lead organizer of a Dagstuhl seminar from September 5th to 10th. The seminar, titled “Structure and Learning,” aims to develop a roadmap that integratively brings together structure and learning in the fields of Machine Learning (ML) and Artificial Intelligence (AI). The hybrid event with 40 international participants was organized together with interdisciplinary experts from all over the world.
The internationally renowned Dagstuhl Seminar series offers researchers in computer science and related disciplines a forum for exchange and collaboration on current and relevant issues in the field. Together with researchers Prof. Dr. Achim Rettinger (University of Trier), Prof. Dr. Jie Tang (Tsinghua University), Prof. Dr. Barbara Tversky (Columbia University) and Prof. Dr. Frank van Harmelen (VU University Amsterdam), ML2R’s Tiansi Dong is organizing the hybrid Dagstuhl Seminar “Structure and Learning” from September 5th to 10th.
Dagstuhl Seminar focuses on the Integration of Symbolic and Neural AI
The seminar combines expertise from the domains of computer science, cognitive science and humor research. It addresses the merging of the two central paradigms of Artificial Intelligence: structure (symbolic AI) and learning, which is traditionally based in neural networks (neural AI). The goal is to advance the development of the “third generation of Artificial Intelligence” that combines the strengths of both paradigms. AI technologies will hence be made more explainable, reliable, robust, rigorous, secure and trustworthy.
Continuous vector embeddings and discrete symbolic rules are images seen from the two traditional eyes of AI. Our goal is to unify the two approaches in the mind of AI and to computationally embody the way of thinking – from the basic neural syllogistic reasoning, to advanced humor understanding. This seminar gathers interdisciplinary experts from around the world to discuss a roadmap about this new style of AI.
– Dr. Tiansi Dong (ML2R scientist)
Around 40 international participants discuss the central topics also in regard to concrete application examples from the fields of AI and computer science, for example Natural Language Understanding (NLU), the Internet of Things and robotics, as well as interdisciplinary application fields, such as humor research. Further information on the Dagstuhl Seminar: Structure and Learning
Tiansi Dong presents his research in the ML2R Virtual Showroom
Tiansi Dong’s innovative research will also be part of the ML2R Virtual Showroom, which is currently being prepared and in which scientific projects and findings of the competence center will be explained in an illustrative and comprehensive manner. With his virtual demonstrator, the scientist will present a method that models human understanding of humor in the context of Artificial Intelligence by configuring ball-like spheres.
TU Dortmund University