Recap of the ICAI Spotlight: Discovery Lab
On October 18th, Elsevier in Amsterdam hosted an event to discuss the work done to date within the scope of the Discovery Lab, an ICAI Lab that studies technology, infrastructure and methods to develop intelligent services for researchers, focusing on finding and interpreting scientific literature, to formulate hypotheses and interpret data. The well-attended event featured talks from each of the ICAI members (Elsevier, the VU and the UvA) and reported on the successful first four years of the lab, which kicked off in 2020.
The afternoon started with a presentation by Michael Cochez and Rinke Hoekstra, highlighting the overall setup of the lab. It has a unique ‘spike’ structure, where Elsevier departments propose use cases for short-term ‘spike’ projects, and the lab directors (Paul Groth, Maarten de Rijke, Frank van Harmelen and George Tsatsaronis) select which of these use cases are most interesting from a research perspective. In each of the 12-18 month projects there are clear corporate deliverables, and data scientists from the publishing company work closely with the Discovery Lab researchers towards an outcome that delivers scientific progress as well as benefits for the company.
Next, five of the researchers involved with the Discovery Lab presented their work: Thom Pijnenburgh from the Elsevier side, and Daniel Daza, Romana Pernisch and Dimitrios Anavalistos from the researchers’ side. Lastly, Anita de Waard concluded with a presentation of the benefits of the lab for Elsevier, which include not just involvement with the state-of-the-art research that the VU and UvA are engaged with, but also offers internal benefits as Elsevier employees interested in innovation meet each other, although they are working in different departments and towards different goals.
After the break two visions of the future were presented by George Tsatsaronis, head of Elsevier’s Data Science unit, and Frank van Harmelen, Professor of AI at the VU. Although not agreed on ahead of time, there was a remarkable overlap in these two enticing visions of the future: both Van Harmelen and Tsatsaronis emphasized the impact of AI on scholarly publishing and the expectation that future research will be conducted in a hybrid setting, which AI supports all aspects of the research process. In the lively discussion that followed many of the participants shared their thoughts and concerns around the involvement of AI into all aspects of the research process. The discussion continued well into the social session after the talks, and was more evidence of the importance of cross-organisational, in-person events to weave the fabric of solid public-private partnerships.