More on user-created content and the wisdom of crowds.

Slate has an interesting article on Why Yahoo! Answers is a librarian’s worst nightmare.

They make some good points about the kinds of questions that are best asked of other people, rather than of a search engine–it’s a distinction I haven’t seen librarians talk about much in our conversations about IM reference and Google.  They also point out how deeply unwise the crowd can be.

I’ve used both Yahoo! Answers and the Amazon.com-sponsored equivalent, Askville, to ask both “facty” and “opiniony” questions.  The quality of the answers ranged considerably–I got some good conversation about the Kindle, for instance.  I also got some pretty lame responses about how vaccines work. I also got an excellent, thorough response about vaccines from a medical librarian who answers a lot of questions in Askville.

Clearly there’s a range of expertise and knowledge in the crowd, and clearly companies like Yahoo! and Google are starting to pick up on that.  Again, it makes me think that libraries should be doing more in this department–leveraging the social participation and knowledge of our extremely smart, educated users, and providing some sort of vetting/endorsement/ratings tool to help people tell the good from the bad from the ugly.

Keeping up with the Knolses

Google recently announced a new venture to help people share knowledge and expertise online, leading with the announcement of their concept of the “knol.”  (A knol is a unit of knowledge.)  Interestingly, their approach here looks a lot more similar to how we might approach this kind of project in academic libraries than it usually does.  They’re concerned with allowing experts to share knowledge and retain their identity as the author of that knowledge, on the assumption that knowing authorship allows users to determine the relative validity and usefulness of information.

This reminds me of the Connexions project at Rice, which uses “endorsement lenses” to allow experts to highlight and endorse educational tools contributed to a common pool.  It’s a new take on the idea of user-created content and the wisdom of crowds:  rather than rely just on popularity as a gauge of quality, these tools acknowledge that some users are more skilled and knowledgeable than others.  It strikes me that this is a compromise approach to content sharing that academic libraries can probably get behind.

The Google “knol” project is still in its earliest stages, and currently by invitation only, but here’s an interesting snippet for academic librarians to consider:

A knol on a particular topic is meant to be the first thing someone who searches for this topic for the first time will want to read. The goal is for knols to cover all topics, from scientific concepts, to medical information, from geographical and historical, to entertainment, from product information, to how-to-fix-it instructions.

Sounds a lot like something that undergraduate students will want to use, doesn’t it?  Sounds like free, authoritative, unlicensed, findable subject research guides.  On every topic under the sun.

I’d be very interested to hear from anyone who’s participated in the project so far.  To me, this looks like something we should be watching very closely.

Public library lending Kindles

The Sparta Public Library in NJ has started lending out Kindles to patrons.  Whether or not the Kindle is the e-book technology to watch, I love this library’s go-get-em attitude, and their creativity in overcoming the issues of accounts, downloading, expenses, and permissions.  A great example for every library that takes a “wait and see” or (worse) a “we could never do that!” attitude.

Dept. of shameless self-promotion: Library instruction tutorials

Friend, colleague, and erstwhile blog reader Annie Zeidman-Karpinski asked what ever happened to those library instruction tutorials I’ve been working on for so long.  Good point!  Here’s what’s new (all tutorials require the free Flash player):

Connecting from off campus (2 min, no audio):  Demonstrates how to set up the proxy server to use licensed databases from off campus.

Find full-text articles using Melvyl (5 min, audio):  Shows how to find online articles if you already have the full citation, by searching for the journal in Melvyl (the UC-wide online library catalog.)

Use Library of Congress subject headings in Melvyl (5 min, audio):  Gives tips on how to do a faster, more efficient search in Melvyl by using LC subject headings.

Search using Worldcat.org (3 min, audio):  Shows how to use Worldcat.org, the free online catalog to libraries around the world, to expand your search beyond the UC system.  (This one’s by Anne-Marie Basso, UCB reference librarian.)

Borrow items using the Interlibrary Borrowing Service (2 min, audio):  Demonstrates how to borrow items from other libraries once you’ve discovered where they are.   (This one’s by Anne-Marie Basso, UCB reference librarian.)

Evaluate scholarly content online (5 min, no audio):  Offers some tips on how to tell whether the article you’ve found online is scholarly or not.

All are created using Macromedia Captivate.  Please give them a test-drive, and send me comments about how they work for you.  If you use a screen reader or other assistive technology,  I’m particularly interested in knowing how they run for you.  All comments are welcome!

Sakai Conference: Google Analytics & Help Documentation

Note:  I had to leave this one early to catch my plane…

Steve Lonn, Margaret Wagner, UM

UM has large Sakai implementation: 18,000 users per day; 3,700 course sites so far

Help documentation is heavily customized at UM; combination of original and modified Sakai documentation, with lots of custom pages for the Ctools implementation. Why collect data on help documentation? It helps to identify tools that have problems, to improve training & support, as well as direct development & improvement efforts, and finally to improve Help docs themselves.

Using Google Analytics to get data on help docs. Started with simple hit counter, but moved to GA to get more information & accuracy.

Screenshots on interpreting GA reports, demonstrating the kinds of information available from the tool

Sakai Conference: Sakai Webcast System

Mara Hancock & Judy Stern, UC Berkeley

Q:  How many institutions have systems in place that automatically capture webcasts/podcasts of lectures? A:  Of those present, only UCB, Michigan, U of Amsterdam.

@UCB, webcasting started in 1995, at the Berkeley Multimedia Research Center (faculty genesis)

2001 ETS formed, and merged with BMRC and media services

2002-2004 ETS updates webcast.berkeley.edu, adds events and admin application

2004-2006 Experiments with captioning and search. Adds podcasting. Delivery partnership wtih Apple and Google. Webcast is next-generation vision.

2007 Strategy, design 7 specs for Webcast next-generation. Development & outreach.

2008+ Open source requirements. Complete Phase 1 development. Planning for Phase 3 and WC learning tools.

Current state of affairs: 6 webcast- and 14 podcast-enabled General Assignment Classrooms. Recharge for special events & courses. Departments share cost of webcasting with campus. All output freely available online to anyone in world. 83 full courses in 2007. 10,000 hours of content since 2001.

Local & world audience: 2 million visits to webcast.berkeley.edu (RealPlayer.) 2.3 million downloads from iTunes U. Youtube 1.6 million views, 8,700 current subscribers. Wanted to put the content up in locations where users could find it (iTunes, YouTube.)

Current questions:  How to make access easier, content more discoverable, share with other schools, make content and context more meaningful (i.e. surrounded by more learning activities in Sakai etc.), leverage innovation from other universities and community source projects?

Needs and barriers: students want more (incoming freshmen consider podcasting essential service), students & public want more portability, quality, quantity. Growing to 70 tech-enabled rooms in 5 years. Approx 1000 courses eligible, 1/3 of undergrad catalog. Problems: patchwork quilt programming, scalable vendor solutions expensive, proprietary code & limited integration, need solid foundation with flexible toolset, other systems not built to play.

UCB is interested in finding partners to join efforts and contribute to or implement OpenCast on their campus. OC Community: a community centered around open, scalable, and sustainable podcast/webcast solutions and best practices for higher education. List: podcast@lists.berkeley.edu. Wiki: confluence.media.berkeley.edu/confluence/x/AoEl Would like to leverage the Sakai community (governance, licensing, etc.) as much as possible.