What is a discovery layer?

Storytime. Last year a few of us from UTS Library were invited to go and talk with students and then help to assess their augmented reality (AR) application concepts that could be applied to the UTS campus. They were advanced Visual Communications students working with some pretty clever and inspiring academics, including a visiting lecturer (Dr Keir Winesmith) who is normally the technical lead for SBS Digital Media. From memory, nearly all of the concepts they came up with were influenced in some way by the students’ use of and experience with social media. One of the concepts was tightly focussed on the Library and based on mobile service including mobile search and discovery and mobile check out.

The student library application included many features that they expected to see and use to search and discover our library’s collections: a basic item record; tags; ratings; reviews; comments; AND the item’s history of use. The history was represented graphically to show frequency and periods of use and even whether the item had been the subject of a fine for late return. We have taken their suggestions very seriously and it has confirmed our belief that we needed to add a basic social media layer to our “discovery layer” with new features such as folksonomic tags, ratings and reviews or comments. We are also looking into the feasibility of adding the item’s history of use.

This experience started me thinking about a number of things. Are we really offering true “discovery”, i.e. the chance of uncovering something accidentally or serendipitously that you may not have been specifically searching for in our online search interfaces? I don’t think so, not yet. They are mostly enhanced search, federated search or unified index based searching. Are we offering our clients, or users, or readers (or whatever you want me to call them Kathryn!) the kinds of services they are expecting to find online now based on their use of social media and various online services and applications that enable profile sharing and which deliver a more personal or shared experience online? No again I’m afraid. To do that I think we need to find out what our clients are doing, observe their behaviours and also talk to people from outside the library world to find out how we might leap ahead of what the predictable, slow-moving crowd that sells us library management systems and so-called discovery layers has to offer. In short, we need to stop walking like Egyptians and learn some new dance steps.

Now, in case you still don’t get it, here are some suggestions that might lead to enhanced serendipitous discovery. They are taken from my own experience with social media and other online services that I think are a long way ahead of our offerings. They enhance your ability to discover new things accidentally through your network of contacts or friends or through the “muddy foot prints” of others who have gone before you and altruistically shared their experience. For me I think it all comes from understanding the power of connections and sharing that is now offered by the web.

For a start, we definitely must start offering these features for our catalogues and search layers: comments (e.g. Flickr); folksonomic tagging (Flickr, Twitter); easy to use ratings (iTunes, LibraryThing); virtual browsing using Cover Flow (I know some libraries are already offering this); and reviews (Amazon, Expedia, iTunes Store).

And now a listing of the other features I like to use and would like to see some of us playing with:

• Little icons that quickly allow you to share a link to what you are viewing on Twitter, Tumblr, Facebook or a blog (Flickr does this very well now). I was also going to suggest little icons to social bookmarking services like Delicious and Diigo, but maybe the toolbar icons that are now added so easily in browsers like Google’s Chrome account for that?

• “Like” icons (Facebook, Tumblr).

• Reblog, retweet or re-post options (Twitter and Tumblr again, and yes Kathryn, I think we have much to learn from the pr0n industry online).

• The optional ability to establish, customise and share online profiles (last.fm) for “your library” that then facilitates the use of favourites (Flickr), “following” (Tumblr, Twitter), asking (Tumblr, Twitter) and things like wishlists (Amazon) which for libraries could mean things like planned reading lists stored for later and shared with friends or colleagues. I see this kind of thing being really useful in facilitating peer-to-peer help or advice that would be helpful to those using our databases or journals.

• Online profiles would also enable features like “scrobbling” your reading, use, borrowing history (last.fm). These profiles allow us to explore through the eyes of others. It works for music because people can easily find music they might like that is well beyond the boring and repetitive play lists of most radio stations.

• A check-in or currently reading/viewing service that might operate something like FourSquare. So, instead of locating yourself geographically, you are sharing where your headspace currently is in the library.

• Randomly exploring what you already know but have forgotten (Apple's Genius) related items (iTunes Store Genius recommendations).

• "Looking within" or sampling from a catalogue entry (Amazon & iTunes: e.g. listening or getting sample of an e-book before you buy).

• Is anyone offering an “I’m Feeling Lucky” button yet (Google)?

• Item use history (UTS students), including the application of late fees!

• Stumbling (StumbleUpon) another opt-in service that tracks your searching, browsing, use or borrowing history and then feeds you other items you might also find interesting or relevant.

I realise that doing all of the above isn’t feasible, nor would it be wise. We do, however, need to try a few of those features and when we set them up we must make them really easy and simple to use. I’ve probably missed a few things, so please let me know what you think.

- Mal Booth

Mal is from UTS Library and is interested in bikes. He's on the twitters @malbooth & blogs here.