The altmetrics are a set of metrics used to measure those impacts of research that differ from the traditional metrics of scientific production, indicators basically derived from the social web, which are generated from the interactions of users in web spaces with publications generated by researchers. These new metrics reflect the influence of scientific work in the new digital context, where social networks and information are grouped together.
Altmetric Top 100 is a yearly list of the 100 most talked works, and so that, they have generated more discussion among researchers and society at large. This list has been drawn up since 2013, according to the Altmetric Attention Score indicator, which is based on 3 main factors: the number of mentions, the mention's sources of information and authors. The statistics, represented by the "Altmetric Donut", are based on the mentions of an article in social networks (Twitter, Facebook, etc.), mentions in policy documents, in blogs, "readers" in Mendeley, etc. The list can be searched by title, journal/collection, publisher, subject area, institution and type of access; and can also be sorted by the different metrics it provides.
During 2019, the most discussed article has been the pre-print "Few-shot adversarial learning of realistic neural talking head models" from the authors Egor Zakharov, Aliaksandra Shysheya, Egor Burkov, Victor Lempitsky, published in the open access pre-prints repository arXiv.org (published version: ZAKHAROV, Egor, et al. Few-shot adversarial learning of realistic neural talking head models. In Proceedings of the IEEE International Conference on Computer Vision. 2019. p. 9459-9468)
