Two members of the UNSILO development team. Manuel Ciosici and Ludvig Olsen, attended the Springer SciGraph Hack Day (London, Friday, June 23rd). The aim of the day was to make use of the tools available using the Springer SciGraph system. Several of the submissions involved linking people, institutions, and research, and the UNSILO project was no exception. It used data from SciGraph to identify reviewers with common subject interests – potentially solving one of the fundamental challenges in the academic publishing lifecycle. As a publisher, you are given an article on a leading-edge topic. To comply with the requirements of peer review, another expert must be found who can state that the article is a valid piece of scientific research, and do to that requires very specialised expertise. There may only be a handful of people in the world who have knowledge of this specific area covered in the article, but how do you find them? “Just read the article” would seem the logical response, but with many publishers dealing with tens of thousands of articles being submitted annually, it rapidly becomes prohibitively expensive to find reviewers by reading all the abstracts.
Manuel and Ludvig created a “collaboration graph” using the SciGraph data, with connections between people if they have co-authored a paper, and connections with institutions if the person has an affiliation. In that way, the system can discover simple conflicts of interest (previous collaborations that led to co-authored papers, or both author and reviewer attached to the same institution).
To make the exercise simpler, it was restricted to one subject (computer science) and one year (2016), but of course in practice the operation could be extended with sufficient computing power. Although the project was a very simple one, it won third prize at the Hack Day. Even so, mentioned Manuel after the event, it might not be good to set too much store by the ranking of the result – the award of first, second and third places was based on how loud the clapping was when the project was presented to the others, something like the practice described in Ancient Greece when a proposal was accepted (or not) by acclamation – how loudly people clapped. It is pleasant to think that state-of-the-art development tools are judged using a method invented thousands of years ago.