October 13th, 2006
Annette Bailey, Newman Library/Virginia Tech
Godmar Back, Virginia Tech
http://www.libx.org/
Situation/problem/reality with users:
- 2% college students start w/ library website
- 89% start w/ search engines
Possibilities/ideas for solution:
- toolbars (e.g., FASTJack, HALbar)
- bookmarklets
- scripts
- Web localization services
Virginia Tech’s chosen solution:
- “LibX” Firefox browser extension
- Installed on client-side
Problems with LibX solution:
- Each ‘version’ works for one library and one library only
- Does only a limited number of things
What LibX provides:
- Direct access to OPAC(s) and databases
- Direct access to OpenURL resolver
- Localization (supports COinS, includes ‘hints’)
LibX features:
- ‘Autolinking’ pages - LibX function that links citations, ISBNs, etc.
- Place library logo in Amazon pages, google search results pages
- Proxy options
- Logos
LibX today:
- 30 ‘live’ editions (including MIT in August)
- Downloaded and installed over 4,000 times
We can do it too:
- Download sources and configure LibX ourselves; or
- Get them to do it (bottleneck)
- Decision: Who hosts it? Us? Them?
Presenters recommend we:
- Go ‘beta’ first - worry about appearance later
- Volunteer to test ‘edition builder interface’ (down the road)
- Subscribe to mailing list
Posted in Firefox, extensions, google, link resolver | No Comments »
October 13th, 2006
Annette Bailey
Annette developed a Firefox extension that allows users to access their library resources on any page - including Amazon, or external databases. The extension can be customized for individual institutions, and they will assist at Virginia Tech. Download: http://libx.mozdev.org.
Posted in Firefox, extensions | No Comments »
October 13th, 2006
Anne Christensen
The University of Hamburg library wanted to assist users with a natural language tool to replace their FAQ. The project was proposed to get grant money for further development. They developed “Stella”. In their character conception, they decided to avoid using a real librarian, created something discreet yet friendly, Stella became a patron saint with a backstory. Along with prescribed answers to questions, Stella is also capable of various moods/expressions including laughing, explaining, enthusiastic, confused.
Novomind Composer is the software used for Stella. It allows only 3000 rules. Stella is placed in a frame that is moveable. Chat is not a big thing in Germany, but this type of interactive tool is really taking off and catching on in libraries. The knowledge base created for Stella is for sale.
See presentation here.
Posted in chatbot | No Comments »
October 13th, 2006
Anne Christensen
U Hamburg
What?
Why?
- Assist website visitors
- uses natural language
- explain access to ejournals and research databases
- info literacy
How?
- Knowledgebase contains 3,000 rules
- Responses are accomanied by one of 10 ‘moods’
- Hooked-up to library catalogue
Who else?
Pros:
- Usage
- Marketing
- 24/7 availability
- Anonymity
Flaws:
- Not info literacy
- No connection to OPAC
- Serious problems with pop-up blockers
- Requires a great deal of programming of responses (not real AI)
Posted in chatbot | No Comments »
October 13th, 2006
Roy Tennant
Access 2006 Keynote Speaker
- “Last Mile Problem” = getting connected to houses they (utility companies) want to serve
- Libraries (historically) are better at description than delivery
- Systems are ill-suited for online delivery
- “Two-click solution” (Jeremy Frumkin): One click to search, on click to get
Books
- e.g., google books = good - but google books is mostly about generating revenue for the publishers
- e.g., library catalogue = bad - takes to long to find link to fulltext
- Librarians have often asked for the wrong thing from their vendors - part of the problem
- Mirlyn mirlyn.lib.umich.edu U Mich library catalog - linking to google books doesn’t work because there’s no search string (no terms, just a link to the book ID)
- Open WorldCat has Schooler/Scholar level-of-audience indicator for digitized books
- FRBR concepts
- xISBN service FRBR relationship
Journals
- Why do link resolvers require users to click to link to fulltext? Why don’t we take them directly there? We could automatically link users to 1st link to fulltext. We could automatically link them to the OPAC (having 1st checked whether or not we have it in the OPAC). We could automatically link them to an ILL form. “Let’s get rid of the window” (and the extra click)
- See Georgia Tech: umlaut.library.gatech.edu/resolve?res_id=523 - uses OCLC’s OpenURL Resolver Registry
- See Dave Lindahl’s GUF (Getting Users to Full-Text):
- floorplans with book highlighted
- uses Cold Fusion, database of print journals, web services API to OpenURL resolver, OCLC’s xISBN service, ISSN lookup, etc.
Posted in Open WorldCat, OpenURL resolver registry, google, link resolver | No Comments »
October 13th, 2006
Roy Tennant
Electronically speaking, libraries have dealt better with description than delivery. As Jeremy Framkin says, we need the “two-click solution”.
It’s not just the vendors faults, libraries have often asked for the wrong things. We need to be better at discovery of online texts. There are a lot of free full text books out there. A good example is the Mirlyn catalogue at the University of Michigan. They link to online texts several different ways.
A direct link into Google Books is confusing since Google expects users to start with a search engine. Directly linking to a book results in a phrase such as “417 references to in this book” - no keyword reference.
The open resolver (ie SFX) is the first step, not the last. Ross Singer’s Umlaut at Georgia Tech is an excellent example of minimizing the number of clicks a user must make to get to the resource. It sucks the data out of the link resolver and puts it up front.
Best linkage from a catalogue, is google-like, linking titles directly to the item.
Posted in Uncategorized | No Comments »
October 13th, 2006
hackfest.kicks-ass.net/
www.flickr.com/photos/moil
www.flickr.com/photos/calvinm
Dale Askley, Kansas State University
- Adding Google books as an SFX target
- plus a catalogue mash-up
Posted in OPAC, google books, link resolver, mashup | No Comments »
October 12th, 2006
Clare MacKeigan (Relais International), (working with Gail Wanner, SirsiDynix)
- “[Patrons] have their tools for finding… now we have to help them get it.”
- future of ILL?
- “…plans to develop an extensible open source browser plug-in and an initial set of extensions based on existing standards. This plug-in will enable people who search web-based resources [to link from citation-bearing webpages to] items from their own library, its resources sharing partners and other suppliers.”
- plug-in must be installed on workstation?
- must gather info on user (occurs when the plug-in is installed)
- there followed a call for volunteers to code an open-source prototype
Posted in ILL, plug-in | No Comments »
October 12th, 2006
Walter Lewis, Halton Hills Public Library
- Used Mr. Sid from Lizard Tech
- Allowed display/magnification of large images
- uses AJAX
- “It works with IE and Mac unless you use IE 5.5 for the Mac which reacts badly”
- applications for GIS?
Posted in image magnification | 2 Comments »