So when I add a paper either from the ACM Library, or IEEE Explore, that I’m viewing through the University EzProxy, Paperpile gets zero details about the paper from the page. I only just realized it works without the proxy. So I assume you’re doing a domain match for feature extraction.
Any chance of changing that match to a sub-string anchored at the start of the URL? The URLs from EzProxy are of the form http://delivery.acm.org.wwwproxy1.library.unsw.edu.au so it seems like it should be doable? And would be nice
I’m a little surprised it got so little though (seriously. It’s basically blank apart from the PDF). These pages still have DOI’s, ISBN’s, and usually BibTex links I thought would be picked up as a general heuristic.