I was interested to see that there is now a page on the Rubin website - Papers citing Rubin Observatory | Rubin Observatory - that provides links to public ADS libraries of scientific and technical papers citing Rubin Observatory. Can you share the queries that you run on ADS to generate those? - maybe they could be posted online somewhere, so that people know exactly what those libraries represent.
I’m interested in this because we need to report on publications to our funders in the UK, so I have been playing with an ADS library myself, for Rubin/LSST papers with UK authors, and it would be good to know how the criteria you’re using compares with those that I’ve used.
There is no query. For the technical papers I did lots of different searches to find older papers and then, since I am on the publications board, I know when new papers are turning up. The ADS library is a curated thing built up from queries and knowledge.
For the scientific papers this is somewhat new in that currently I’ve been doing it manually (with some help from ADS queries such as full:"LSST DP1") but it can’t be entirely automated from queries because some papers will cite the DP1 paper or mention DP1 in passing but not formally be papers working with DP1 (if you write a paper on Euclid and mention DP1 in passing, that is not a Rubin paper). In theory, citing the DP1 dataset DOI could be a proxy for a Rubin DP1 paper but we cannot guarantee that people will use it consistently.
I am currently working on handing over the scientific publication tracking to CST and working with @MelissaGraham and @ChristinaAdair on the methodology. NOIRLab has some experience with this through Sharon Hunt (the librarian). There is talk of writing a tech note to fully define what it means to be a Rubin science paper. Maybe you could help with that work by giving the UK perspective.
At the moment, we’re waiting to finalise the larger list of “impact metrics” that STFC want us to report on, and I have been hoping that, for this, they would be satisfied with something that could be generated with a query, rather than requiring significant manual effort - although, as you say, there would be an inevitable loss of quality in doing so.
(p.s. actually, thinking further, we could just take your library and run an additional filter for authors with UK affiliations…)
If you can do the UK affiliations query then I’m pretty sure you can combine that with the ADS library because you can always convert the library few to a ADS query.