I can come to your library, book club meeting, or conference to talk about how to help your readers find their next good read. Click here for more information.

Wednesday, July 15, 2009

ALA Report- From the Book and Beyond: Interdisciplinary Readers' Advisory

Beginning with this year's annual conference, RUSA will now have an annual RA Research and Trends Forum. This year it debuted as the RUSA President's Program. If this program was an indication, we will be very lucky to have such a wonderful RA specific program offered at every conference from now on.

This year's program featured three speakers and was entitled "From the Book and Beyond: Interdisciplinary Readers' Advisory." Here is the link with more details.

First up was Nathan Altice, a musician , digital artist, and adjunct professor of sound communication at Virginia Commonwealth University. Altice talked about the brain, technology, and musical appeal. He focused his talk on Pandora, the for-profit (but free to end users) Internet radio station where users can create customized play lists based on computer generated recommendations.

Pandora makes it recommendations based on the "Music Genome Project." A group of musicians and music professors got together and analyzed thousands of songs, creating 400 plus appeal terms (click here to see them all). The founders of Pandora then created a proprietary algorithm that matches the appeal terms with certain songs; creating "listen alikes."

I was familiar with with Pandora, having used it and been amazed at how well it does, but I had no idea how it worked until now. Altice's main point was that this works only when people and machine work together. To paraphrase him, humans are best at assigning meaning and the machines can do the computations, "the digital muscle needs human intervention."

Altice's program made me think about how we make suggestions to readers. Generally RAs shy away from computer driven recommendation engines. But ones for books like Gnooks, do serve their purpose at times. For example, I use Gnooks when I have no idea where on the literature map an author is. Click here for a literature map on Charlie Huston to see what I mean. The point is, there is a human/machine interaction that makes these recommendation engines at least worth taking a look at. Also, I found the discussion of how Pandora breaks down appeal and then applies it in a mathematical way very intriguing.

From music, the discussion moved on to Steve: The Museum Social Tagging Project and a presentation by its founder Susan Chun. Basically, this project has taken many of our country's museum's digital collections and opened them up for the public to describe them. Chun, who at the project's start was at the Metropolitan Museum of Art in NYC, was noticing there was a huge semantic gap between the catalog records and how regular people describe (and as a result) search for materials.

T0 make a long story short they wanted to see if allowing social tagging would help users find art more quickly, change the ways users engage with the museum, and help museum curators understand what their visitors see? So they let people start assigning words to their pictures online. They found that 88% of the users tags were relevant. But now they are trying to sort out the "noise" of too many tags. They are also working on a way to determine whether or not a specific tag is useful.

I was keenly interested in Chun's findings because in our library system, we have begun the switch to a new interface for our catalog (Encore) in which library patrons will be able to start tagging books. The hopes are for it to have the same success as Steve, replacing outdated subject headings like cookery with the more likely term, "cooking," to be tagged by end users. I will continue to follow the Steve Project as we get closer to beginning our own system wide social tagging project.

Finally, Nora Rawlinson from Early Word: The Publisher Librarian Connection came on to talk about the results of a publishing industry study (the Codex Study) as to what makes people want to pick up a particular book. Notice she didn't say what made them buy it. The publishers first need people to pick up their book before they can move toward making a purchasing decision. She then took each reason and talked about how libraries can use that information.

The top reason (63%) was that it was by a favorite author. Libraries have this covered with our standing order author lists and our extra copy plans. But, Rawlinson did mention that this should remind us to always ask readers who their favorite authors are.

Second place was a tie between the Genre of the book (48%) and the Flap Copy (48%). Genre is no shock to librarians. We all have genre collections, but the Flap Copy importance was big news. Usually publishers give that job to the lowest person on the totem pole, but now they are starting to take flap copy more seriously. She even gave us a few examples of flap copies that are beginning to contain readalikes (for fans of...). Librarians should learn from this more journalistic writing style in selling books to readers.

The findings rounded out with 29.4% going for Cover Art and Design and 23% who were drawn to the book's title. (Obviously people could pick more than one thing.)

Rawlinson wrapped up with a few final thoughts. When we are buying for the library should we consider book jackets, and how? Libraries could use flap copy for shelf talkers to draw out books lost in the stacks. She also suggested that librarians read some of the ugly covered books to see if they are worth reading. If they are, we could jazz up the covers or put shelf talkers right on the books themselves.

As you can see it was a very thought provoking program. As an experienced Readers' Advisor, I liked how this program made me think outside of the box. Can't wait for DC next June.

1 comment:

ckubala said...

Thank you Becky for this informative report. Gnooks is a new one for me. It looks like it has potential. I'm going to use it with one of our fussy readers who is always looking for new authors.