notes & links

LibX (link to podcast)
Annette Bailey - Virginia Tech
(and Godmar Back, Computer Science @ VA Tech - he didn’t present, but was a partner in the project)

More information available at: http://www.libx.org/

I was interested to see this talk, because LibX is one of those things I’ve seen referenced all over the Internet, but haven’t had a chance to really look at or play with myself. The talk-and-then-demonstration structure of most of these technology talks was really helpful. The tool itself was very cool. After highlighting text on any webpage, the user has a variety of search options available to them. Links to library resources can also be included in search result lists, or on Amazon pages.

The overwhelming question I had, however, at the end of the talk was “how do you get your users to install the extension.” That question didn’t get answered immediately, but it came up more than once as the conference continued. In fact, in his multi-topic Thunder Talk, Dan Chudnov argued that applications like LibX, that users have to install, aren’t the right direction for improving the users’ experience. I was also interested to hear that this tool was developed on the developers’ own time, on a short deadline, and that this is why it was developed for Firefox. That part itself wasn’t all that interesting, but the fact that the VA Tech library doesn’t have Firefox on its computers, so this tool isn’t useful to students working in the building at the school for which it was created, was. As a matter of fact, one thing I found striking throughout the conference was how often people were most excited to talk about things that they were working on on their own time, or in their spare time. That suggesed a couple of issues - first the need for library work environments that support innovation and change, and also the extent to which there is a potentially awesome developer community in libraries being shut out of working on the systems that control much of what we do - they’re working around the ILS’s, not with them.

That said, I do think that library staff and librarians would find LibX excessively useful - especially those staff in acquisitions or ILL who have to spend a lot of time repeating searches between search engines and library tools. And the way the user numbers skyrocketed after MIT went live with their LibX version suggests that there are users out there motivated enough to install an extension.

Improving the Catalog Interface with Endeca - Tito Sierra (NCSU) (link to podcast)

Additional Information available at http://www.lib.ncsu.edu/endeca/

This was the first talk where we heard the word Endeca, and by far the most comprehensive discussion of its capabilities. Tito Sierra said at the start that this talk would focus on the discovery portion of NCSU’s catalog. Their goals were to improve the user experience and make the MARC data work harder. They chose Endeca specifically because there were lots of commercial examples to look at, and they thought it would improve the users’ relevance rankings, improve their browsing experiences (both faceted browsing and true browsing), and give them faster searches.

This talk was especially useful coming so early in the conference because it provided a quick, visual introduction to a lot of the major concepts that would recur throughout the two days: faceted browsing, search “comforts,” the interactions between the presentation interface and the backend, etc.

Better relevance, better speed, a locally built and managed presentation interface and persistent parameter based encrypts (which allow persistent links to searches that can be embedded elsewhere) were described as postiive outcomes. Less positive were some things not supported by Endeca: personalization, folksonomies, etc.

This talk was also interesting to me in its specific discussion of the process of building and launching the product - Sierra argued for small teams and launching before the product is “done.” The team responsible for this project had only 7 people, with the charge to launch something when it was good and then improve it after launch. These two factors seem to be necessary to creating the kind of nimble and flexible organization that can stay ahead of changes to the information technology landscape, and they might suggest why NCSU so frequently is out in front of the curve with trying new things — it can’t be that they have better resources than everyone else. Last year at ILI one of the most worthwhile talks I saw was also from an NCSU librarian who argued for a similar approach - small teams, and trying new approaches (which sometimes don’t work - another thing that can be difficult in many libraries).