At some point in my graduate career, some kind person told me that I should be subscribing to arXiv emails. This was, of course, a big improvement over not paying attention to what was on the arXiv at all, but still more trouble than it should be. the arXiv classification is way too crude for my purposes, and I just get overwhelmed.

What would be much better is if I could get emails compiling submissions of particular authors (effectively, arXiv-stalking them). Anyone have ideas on how to do this?

EDIT: I ended up writing a Yahoo Pipe that does this. You can use it yourself here, and if you clone it, you can use your own list instead of mine. I’ve also included some keyword sorting, to address the concerns of David and Allen.

10 thoughts on “arXiv-stalking

  1. Some people use services such as Yahoo Pipes to filter the arXiv RSS feeds into two smaller feeds based on abstracts having certain keywords of interest.

    A site that permits sorting articles based on the co-authorship network is

    Cosmo-coffee allows author’s names into mix as well as abstract keywords.
    (Although the discussion fora focus on physics, they do have an index of all math articles for the purposes of your author filter)

  2. My low-tech “solution” is to bookmark various authors’ search listings, and open in tabs once in a while. I suppose you could automate that by telling cron to trigger wget and feed into a text-matching program or something.

  3. I’d like to urge others not to follow Ben’s model. I think that following subject classes is a much better way to find interesting and important math than following authors. For example, I read the abstracts for Algebraic Geometry, Combinatorics and Representation Theory every morning, and I often find worthwhile work by people I’ve never heard of. If I were to stalk my favorite authors, I doubt I would find almost anything that wasn’t in one of those subjects.

    I like the suggestion of keyword filtering though. I may implement that to save myself some reading.

  4. I have no advice on how to do this, but it sounds like a bad idea. What about when someone you don’t know (e.g. some grad student) posts something relevant to what you do?

    I was overwhelmed by the e-mails until I realized I shouldn’t be subscribed to more than math.SG and math.GT. (In particular, I am not an algebraic geometer and math.AG was dominating my e-mails.) Anything else I’m interested in hopefully will be cross-listed.

    I suppose your issue could be that the subject break-down does not work well for what you work on, in that your work touches on many of the major subject areas. Perhaps a solution is to have the arXiv e-mails automatically grepped for MSC classes that are relevant to you?

  5. I’ll just note: it’s not that I’ve decided this should be the only way I’ll look for papers; it’s just that I’ve noticed I’ve started to miss papers by authors whose papers I know I should read, and am trying to avoid this in the future.

  6. Incidentally, looks cool, but it seems to only be possible to use for astro-ph and cond-mat for the moment, which means it’s basically useless for mathematicians. Thanks a lot, whoever it is that runs arXivsort.

    Also, the website seems to have broken some time between when I first looked at it and now.

  7. Ben,

    If you do keyword filtering, you might want to use to do a google blogsearch on the matter. Once this is done, Google provides you with an e-mail blog alert on the left side of the result screen. Since arxiv is blogged on a livejournal feed, Google will send you an email when new elements with the keyword show up. It will also send other items not in arxiv as well.


  8. I always enjoyed sifting through the 25 sections of arxiv only to find six months later I missed the really interesting one.

Comments are closed.