These are random notes from the sessions I attended Sunday. I need a new laptop so I didn't bring my tired old one to live blog - these are from my scribbled notes on paper.

How much change do you get from 40$ - Erik Choi - this was a typology of failures in social q&a. The system offers some suggestions for how to do better questions but I think their intention was to use this research to help people ask better questions. As Joe Hourclé pointed out in questions - Stack Exchange supports query negotiation/refinement but they're looking at what to do with Yahoo, which is the most popular and has a lot of failed questions. Their big categories were: unclear, complex, inappropriate (prank, awkward...), multiquestion.

Dynamic query suggestions - dynamic search results - Chirag Shah. This was looking at google's way of showing you results as you type and also offering search completions as you type. Google says it saves 2-5s per search, but they wanted to test it. They did it in a laboratory setting with 3 conditions - neither, only autocompletion, all. They gave a task asking users to search for information on the velvet revolution and other revolutions and they looked at the number of pages viewed, concepts (noun phrases?) used, eye tracking. The dynamic stuff didn't change the number of concepts in a query, the queries were shorter but not necessarily better.

How do libraries use social networking software to communicate to users - they looked at big libraries in English-speaking countries and "greater China" (Taiwan + Hong Kong + PRC). They looked at the posts and interviewed a librarian from each. Some discussion afterward how Weibo is better at supporting conversations than Twitter - it would almost have to be

Barriers to collaborative information seeking in organizations - I'll have to read this paper... he spent too much time on methods and really cut his results section short.

This was originally posted on my blog on November 17, 2007. Deborah Barreau passed away this morning from cancer. There's a lovely message from Gary Marchionini on asis-l.

---

This post is a review and commentary on: Barreau, D. (2008). The persistence of behavior and form in the organization of personal information. Journal of the American Society for Information Science and Technology 59, 307-317. DOI: 10.1002/asi.20752

Goal: Barreau re-visits her 1993 study (published in 1995) in which she interviewed seven managers to determine how they manage electronic documents. In particular, in her 1995 study, her goal was to examine how Kwasnik's (1991) dimensions of organization of print materials translated into the electronic domain. In this study, her goal is to learn what has changed in the more than ten years and what impact new technologies have had.

Methods: Her sample consists of 4 of the 7 managers interviewed in her earlier study. She asked the participants broad questions on what personal information they have in their office, how they got it, how they organize it, and how they find things in it. She also asked what changes they would like to see in the technology.

The responses were coded using Kwasnik's dimensions. No information is provided on how the interviews were conducted and how the coding was actually performed. There are mentions of transcripts and notes, however. A sample of the statements were "double-coded" and an intercoder reliability check was done. (I almost missed this bit because the html is a bit goofy to read)

Results: I will just pull out a few interesting points here.

• participants saw their intranet as an extension of personal space when they had bookmarked or used send to desktop as a link to keep some information.
• they bookmark stuff and then never use it
• participants were split between keeping a clean e-main in box by acting on or deleting things immediately and reporting that their e-mail was out out of control
• retrieval is through browsing an ordered list

Changes they would like to see: synchronized single sign on

Conclusions: Many things remained the same. The way the managers name files, and use catch-all directories were two things in particular. Some things that have changed include the extension of the personal space to include bookmarked things from the web and the sheer number of different systems required to do the job. New dimensions are suggested to update Kwasnik's listing.

Commentary: My immediate reaction to this article was very positive -- mostly perhaps because it resonates with my own findings (Pikas, 2007). More information on methods is required to adequately judge the validity and transferability of this work.

She makes the point that corporations need to do better to back up user's work. This is something that also came out in my study. It could be that the corporations *are* doing a good job of backing information up but are not *communicating* well enough so that users trust the backups.

She also makes the point that organizations need to do better with e-mail. First, for records management purposes, they should discourage the retention of older e-mails. I strongly, strongly disagree with this. Much valuable information is included in e-mails - only in e-mails - and there should not be an arbitrary retention policy requiring their deletion if the user finds them useful (yes, I do know about e-discovery, but if you're not doing anything wrong- I guess I'm naive). Second, she states that organizations should do something about advertising e-mails received (ok, this is fine), about broader distribution lists than are required for the job (ok, I was getting e-mails in Maryland once for things lost and found in the Philadelphia office- so this is clearly a management issue), and about too many interruptions. I disagree about the interruptions truly being a something that the organization as a whole can/should fix through rule making. This article speaks to me that more training is required on the effective use of e-mail and IM. Perhaps the users should employ a do not disturb message on IM and log out of e-mail if they are working on an intensive task.

This is my first use of the BPR3 logo so I would be happy to take comments on that (or complaints if I'm not doing it right!)

---
Barreau, D.K. (1995). Context as a factor in personal information management systems. Journal of the American Society for Information Science and Technology , 46(5), 327-339. DOI:10.1002/(SICI)1097-4571(199506)46:5<327::aid-asi4>3.0.CO;2-C

Kwasnik, B. H. (1991). The importance of factors that are not document attributes in the organization of personal documents. Journal of Documentation, 47(4), 389-398.

Pikas, C. K. (2007). Personal information management strategies and tactics used by senior engineers. Proceedings of the Annual Meeting of the American Society for Information Science and Technology, Milwaukee, WI , 44 paper 14. (This will be made available open access 90 days after the conference)

Labels: PIM

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.

Ever since MPOW got rid of our physical collection, I’ve had more and more to do with ebooks. We can’t replace the collection we built over the course of 60 years but we do have access to a lot of stuff. The biggest single provider we have is Springer for many reasons. Their books come as unlocked pdfs, one per chapter. So they kind of in a way work like journal articles. I mean, you’re on a platform that also has journals and it really doesn’t seem much different. But…. if you’re presented with a print book and you need to find information in it, you rely on the training you had from librarians and teachers from elementary school onward. You start with the table of contents or the index. Makes sense for an ebook, too, in many cases. People instead recognize it as a web search pattern, not a book.

I have a hard time convincing our technical staff to try the ebooks. They are skeptical. They don’t believe they’re actually getting the book. They don’t think it’s going to work.

So it was with interest that I saw this evidence summary in the new issue of Evidence Based Library and Information Practice: Undergraduate Science Students are Uncertain of How to Find Facts in E-books Compared to Print Books by Christina E. Carter.The students thought the ebooks should work like Google but they didn’t. The reviewer commented that they didn’t compare the ebooks to journals, which I pointed out above.

This was reviewing: Berg, S. A., Hoffmann, K., & Dawson, D. (2010). Not on the same page: Undergraduates' information retrieval in electronic and print books. The Journal of Academic Librarianship, 36(6), 518-525.

I’m not going to summarize the summary further – particularly since it’s open access.

I take it for granted about the standard ebooks (not the goofy platforms that require some bizarre check out and download and authentication), but I shouldn’t. People still don’t know how to use them or even believe they exist.