Single Figure Publications is an interesting idea by William Mobley published in F1000Research. F1000Research is “a publishing platform offering immediate publication of posters, slides and articles with no editorial bias, [but with] transparent peer review” Mr Mobley proposes in his editorial that Single Figure Publications (SFP) should be a new format of short text pieces, shorter than a scholarly article, and, tantalisingly, close to machine readable. What does he mean by this?
On looking at the Mobley article, actually an editorial, it is maddeningly vague. William Mobley states that SFP is a "novel, efficient format by which to communicate scholarly advances", which is fine, but then goes on to describe it as "a forerunner of machine-readable nano-publications as defined by the W3C". Now, nano-publications is a W3C idea for machine-readable text (see, e.g. "The anatomy of a nano-publication"at the W3C website). In other words, he is proposing much more than just short scholarly articles; he is proposing machine-readable articles. But without any examples it is difficult if not impossible to establish what he means by a forerunner. If you want to create scholarly text in machine-readable form, such as triples, the onus is on you to state how you will author such triples so they are both human-compiled and machine-readable. I checked with Dietrich Rebholz-Schuhmann, a leading figure in computational linguistics at the University of Zurich, who stated no publisher is creating nano-publications. They might be a great idea, but until they are realised, they remain just an idea. In the meantime, extracting machine-readable statements continues to be done via manual curation or by text mining.