It’s not programming, it’s not in my job description, but it took a lot of time these past 2 weeks

Aug 07 2011 Published by under Information Science

I didn’t participate the library day exercise this time – that’s when you document a day in the life in your job as a librarian so that others can see what it’s like to be a librarian.  I have a somewhat non-traditional job so my normal day in the life isn’t like the normal day in the life for most other librarians. It struck me earlier this week – as I got frustrated with things going wrong on 3 fronts at once – that I’m not even sure how to describe the work I’m doing.

Project one: we have an intranet search service that goes way beyond the standard appliance you might purchase. In addition to connecting to our document repository, sucking in the index from our web crawler, and indexing our SharePoint installation, it has an “expertise” tab. This is an index of custom compiled profiles for our employees with information pulled from their resumes, their MySites (a part of SharePoint where you can have a profile), the corporate directories, and, more recently, internal grant submissions and social network participation. One of the obvious things that’s missing is a listing of the external research articles the employees have written. For various complicated reasons, another librarian and I maintain the most complete listing of these documents.  We take alerts from all of our various databases, import the records into RefWorks, and then export them in a custom built Movable Type format which we then import into our listing which is on a MT blog. The tool we’re using can’t just index the web listing, as the author names are not pulled out and we need to link the articles to the directory IDs so that they turn up in the right profiles. The obvious thing was to just export the RefWorks records in some tagged or XML format. The only thing is that this would show the citation, but not provide any help on getting to the full text (we have a very hard time marketing our services). Great, so you can’t update an export format but you can make a custom bibliography with an open URL link. Anyway, to end this long description, I had to completely make an export format from scratch so that another librarian could import it in to Excel, run a script to parse out authors and match them to the directory ID (most of the time), and then upload them to a SharePoint list (ew.)

Project two: my larger institution has been working very hard on migrating to a new interface to our catalog. This runs on Blacklight, an open source effort led by UVa. It’s fabulous and we’re all very excited about it. Unfortunately, this means that other tools that made calls to the old interface will have to be changed. This includes Z39.50 services and things like LibX. If you aren’t aware of LibX, it’s a fabulous browser add-on that adds cues to bookstore websites so you can see if the book you’re viewing is available at your library; hotlinks PubMed IDs, DOIs, ISBNs, ISSNs so that you can click on them to see if the resource is available from your library; lets you reload a page through your proxy server for off-site access; and lets you search for highlighted things on the page in your catalog or other services you’ve added.  Ok, so obviously my larger institution’s LibX needed to be updated. I’m the only one left who knows how (although I should have been training 2 other people but have not  - not their fault, but mine since I’ve been busy) and I’m the primary maintainer. Mostly because I volunteered. Anyhow, I was totally baffled by the edition builder for a while, but then I was able to see what UVa did with theirs and then edit that. Basically I tried to take theirs and then substitute in any changes I knew about and then sent the information to our real programmer (who has been slammed with work) to see if I was close. He made a couple of suggestions and we were off to the races… except for … CRAP! It uses an OCLC service that needed to have our institution’s registry updated. Neither of us knew who was *supposed* to fix that, so I put in a couple of tickets and changed my lab’s registry information while the programmer changed the larger institution’s registry. And crap again, because then it came back thinking we could only ever search one ISBN at a time which is not true – you can OR a large (if not infinite number). Finally, I got that semi-fixed (it now works for up to 25 ISBNs coming back from the xISBN service). We’ve tested it and I pushed it out live on Friday.

Project three: we’re updating our internal portal page for information services – we had planned to suck in listings and descriptions and what not from the services offered by our parent institution… but ARGH, the site is being built in some version of SharePoint and of course it’s acting all wonky. Even embedding a catalog search is turning out to be a hassle.  So I started listing resources and building resource guides – which is actually a very typical job for a librarian… next job is to help the people figure out how to embed these services even though a) I’m not a programmer b) I don’t know anything about SP and c)I have other stuff I need to be doing

Project four: I’m embedded in a team in a sponsor-facing department working on a distributed knowledge management system that’s running on Semantic Media Wiki. Unfortunately the two members of the team with whom I work most closely have been completely pulled off for a few weeks to work on another project so that leaves me. So I’ve learned how to write these hugely complicated nested queries using arraymaps and templates to display results. Once again, it’s not programming, but it’s not literature searching either. It doesn’t seem that difficult … but it was, for me. I wish I could show this off, maybe eventually once it’s delivered.

Project five: It started with standard scientometrics, but I was working with an hci/visualization expert who prototyped a new system for exploring and visualizing connections, etc. I had already done a lot of data cleanup and visualization, but he needed data in a different format. So this is another example of me messaging data export from various tools (Sci2, VantagePoint, and originally WoS) for import to another system.  We also did a proposal for an internal grant for next year to continue working on this so that took some time.

 

I think there were more, but these were the ones that struck me. It’s not programming, it’s really messing with settings with existing products, but that doesn’t seem to really capture the complexity or frustration. Oh, and in the middle of this my institution changed over to default to the new RefWorks interface – good – so I had to redo the tutorials (only got one done). About 18 hours after I did the update and announced it, the interface changed again so I needed to redo a few screenshots… sigh.

One response so far

  • Robin Sinn says:

    Hang in there, Christina!
    Thanks for the description; you've given me a great description of what you do.
    Robin