Title : Re-uniting AI, a new course for the AI Master "Combining Symbolic and Statistical AI"

Presenter Frank van Harmelen
Abstract AI has long suffered from being split in (at least) two schools, the symbolic and the non-symbolic. Until recently, those two schools of thought were barely communicating. In the recent year or two there has been an inspiring wave of attempts to combine the two schools of thought, trying to combine the strong points of both. As part of the joint VU/UvA Master in AI, we (Annette, Sara and me) are teaching a new seminar-course where with a limited number of students we will read the recent literature on combine symbolic and statistical AI. In this WAI presentation I want to outline the course, present some of the literature we will read in the course, and hopefully get inspiration from you on other literature to read and possible student projects to do.

Title : Rethinking Data Publishing

Presenter Tobias Kuhn
Abstract Many approaches have been proposed to deal with the increased importance of datasets in science and beyond. Existing data publishing approaches, however, have almost exclusively been modeled after classical publishing processes, with a publishing authority responsible for ensuring availability and integrity of published objects of the form of datasets. But couldn't there be alternatives to this fundamental publishing model that are better suited for data publishing and have better properties in terms of reliability, scalability, and trust?