Wednesday, July 22, 2015

Altmetrics, Disqus, GBIF, JSTOR, and annotating biodiversity data

JSTOR Logo RGB 60x76Browsing JSTOR's Global Plants database I was struck by the number of comments people have made on individual plant specimens. For example, for the Holotype of Scorodoxylum hartwegianum Nees (K000534285) there is a comment from HÃ¥kan Wittzell that the "Collection number should read 1269 according to Plantae Hartwegianae". In JSTOR the collection number is 1209.

Now, many (if not all) of these specimens will also be in GBIF. Indeed, K000534285 is in GBIF as http://www.gbif.org/occurrence/912442645, also with collection number 1209. A GBIF user will have no idea that there is some doubt about one item of metadata about this specimen.

So, an obvious thing to do would be to make the link between the JSTOR and GBIF records. Implementing this would need so fussing because (sigh) unlike DOIs for articles we don't have agreed upon identifiers for specimens. So we'd need to do some mapping between the specimen barcode K000534285, the JSTOR URL http://plants.jstor.org/stable/10.5555/al.ap.specimen.k000534285, and the GBIF record http://www.gbif.org/occurrence/912442645.

In addition to providing users with more information, it might also be useful in kickstarting annotation on the GBIF site. At the moment GBIF has no mechanism for annotating data, and if it did, then it would have to start from scratch. Imagine that a person visiting occurrence 912442645 sees that it has already attracted attention elsewhere (e.g., JSTOR). They might be encouraged to take part in that conversation (because at least one person cared enough to comment already). Likewise, we could feed annotations on the GBIF site to JSTOR.

A variation on this idea is to think of annotations such as those in the JSTOR database as being analogous to the tweets, blog posts, and bookmarking that altmetric tracks for academic papers. Imagine if we applied the same logic to GBIF and had a way to show users that a specimen has been commented on in JSTOR Plants? Thinking further down the track, we could image adding other sorts of "attention", such as citations by papers, vouchers for DNA sequences, etc.

It would be a fun project to see whether the Disqus API enabled us to create a tool that could match JSTOR Global Plants comments to GBIF occurrences.