[Taxacom] Automated monitoring of the taxonomic literature

[Roderic Page]

Hi all,

Thanks for the feedback (for some reason Stephen’s ended up in my spam filter so I was a bit puzzled by references to his comments).

If I can summarise, the comments are:

1. Coverage: Yes, this is clearly a problem given the widely dispersed nature of the taxonomic literature. I’m hoping to greatly increase coverage, based in part on other work I’m doing to get the taxonomic literature into Wikidata. The sources for BioRSS at the moment are limited and mostly the “obvious” ones. I take Stephen’s point about “obscure” journals, I hope to add more of these shortly. John’s dream of journals publishing their contents with a “tag” would be nice, but there seems little incentive for them to do this.

2. Taxonomic databases as a source: Scott raised this possibility, and it is one I’m looking at. Ironically it was a discussion with Peter Uetz that promoted me to get this project out the door - he is interested in a tool that monitors the literature for new publications on reptiles. The problem with taxonomic databases as a source is that most treat the literature as a second class citizen (no DOIs, for example), plus there is the issue of dates - if I harvest a feed of additions to a database, many of those additions will be older literature recently added, whereas I’m after just the recent literature. But certainly worth exploring.

3. Dates: I take Neal’s point about dates, this gets messy as RSS feeds may report the date the article was published, the date the feed was created, or some mixture of those. I’m trying to resolve this so that the two dates aren’t conflated.

4. Machine readable. Tony makes the point about whether this tool is intended for use by people or machines, to which the answer is both. I think the place machine-readability will shine is as these publications get added to Wikidata, which is where detailed links between publications, people, journals, funders, and taxa are starting to be made.

5. Benchmark. A common these is that for this to be any good it needs to be better than Google Scholar (or at least find things it doesn’t). That’s a high bar, but a good target to aspire too.

Thanks again for the feedback, I’ll see what progress I can make in the coming weeks.

Regards,

Rod