Home page - Speakers - Abstract - Jan Velterop

Abstract - Jan Velterop

Jan Velterop

Concept Web Alliance

(click on the title to see the presentation slides)

"New way of opening up the scientific literature"

What if we could freely get to all the essential knowledge contained in scientific articles, without those articles themselves necessarily being open access? And what if such free access to the essential knowledge is actually in the interest of publishers as well? For that, we need to split the roles of papers as ‘scientific record keeping’ (the ‘interface with officialdom’) on the one hand, and ‘knowledge dissemination’ on the other. This is made possible by what we call “nano-publication”. The essence of scientifically meaningful content, especially in the natural sciences, can to a very large degree – often completely – be expressed in statements such as ‘DMD interacts with SNT1’, which are virtually always ‘triples’ of the form subject > predicate > object. Such triples can be extracted from the literature, and, particularly when semantically disambiguated, lend themselves optimally for machine reading on a large scale and for efficiently conveying knowledge and facilitating analyses. Triples can be seen – and published – as ‘nano-publications’. And they would naturally be open access. Nano-publications could easily be extracted, and where needed validated by the community, if the publishers shouldn’t do it themselves. Whilst in principle nano-publications extracted from classical papers would be subject to copyright, this could in practice only be used to ensure proper acknowledgement. Putting up payment or legal barriers to access would not be tenable. Imagine the information “this statement made by author X published in article Y in journal Z” being put behind tollgates. That would be tantamount to putting the information that “this book was written by author X and is published by publisher A” behind tollgates. Nano-publications that are rich semantic triples are in essence references, and wide and open availability of references to the content they publish is what most publishers crave. Nano-publications are therefore necessarily open access. And this open access is actually beneficial not just to scientists, but to publishers as well, be they open access publishers or traditional ones.

Supporting organization:
French Ministry For Higher Education And ResearchMax Planck SocietyINRIA - French national institute for research in computer science and controlUniversité Paris 1 Panthéon Sorbonne