It’s great that our academic libraries can provide the community with local and remote access to vast amounts of digital content. But in doing so, do our OPACs and aggregator databases eliminate the benefits of the serendipitous discovery of information? We’ve probably all had the experience of coming across an article or book while wandering the stacks or periodicals area. I’ve discovered a good many articles in the education technology literature just from browsing a journal cover that was left by the photocopy machine or waiting to be reshelved. With structured searches in electronic resources we may find what we seek, but the opportunities to come across something in a random way are fairly rare.
I hadn’t thought much about the loss of the serendipity effect until I came across an article titled “In Search of Serendipity” published in the Wall Street Journal. This article also caught my eye because it’s based on the newspaper industry, of interest owing to some of the Internet Age struggles that it shares with the library. According to the author, claims that online newspapers lack serendipity are just not true. With a print newspaper it’s not uncommon to spot an article of interest that you might not normally read. But how can that happen in an online newspaper environment? The author points to a feature of online newspapers that’s becoming more common. The “Most Popular” or “Most Viewed” list. I know I’ve discovered more than a few articles I would have never searched for from these lists.
I got to thinking that this could be a desirable feature for OPACs and aggregator databases that could return an element of serendipity to the library research experience. Why not have a box that appears in your search screen that identifies the top ten articles or books retrieved by similar searches to the one being conducted. That’s not quite like the “more like this” function in library databses which is just another, but somewhat more refined, search. Instead it would draw upon what others have searched recently that is similar in nature to your own search – and quite possibly just show what others found interested to actually view. Perhaps the box would simply show the ten most recently viewed articles from the past 24 hours. The author does note some differences between “most popular” and “most e-mailed” articles for delivering serendipity online.
As our academic libraries grow increasingly digital we will be removing opportunities for old-style serendipity. Now is a good time to start thinking about ways in which we can inject the value of serendipitous discovery into our research resources.
That’s a great idea. Other possibilities include showing the items with adjacent class numbers (the equivelent of what’s near it on the shelf!), showing items _not_ matching your query, but sharing some of the same LCSH headings as the ones in your results, etc. etc.
There are lots of ideas like this, and in truth, I think they will make the interface better at helping the user find what she was in fact looking for (but didn’t know how to enter the ‘right’ query for), not just help her find what she didn’t realize she was looking for at all (the implications of ‘serendipity’).
This is one thing I like about the WPOpac: http://www.plymouth.edu/library/opac/
It would also be great to display random book-cover images (or citations from a federated search) from the collection each time a patron visits.
On the topic of serendipity, I came across this article a month or two ago by William McKeen in the St. Petersburg Times:
http://www.sptimes.com/2006/03/26/Perspective/The_endangered_joy_of.shtml
I want to chime in and say that it’s been my experience that my blog aggregator is a great serendipity facilitator. Almost every day, as I click on someone’s blog, I’m led to another blog of interest. In fact, for me, that’s one of the main attractions of blogging. It’s a habit forming serendipitous experience!
I’ve never accepted the notion that digital resources are intrinsically less conducive to serendipity that print collections. In fact, I’d say that for most users, spontaneous and random discovery occurs more readily in a digital environment. Most aggregated periodicals databases, for example, offer hyperlinking not only to “more of the same,” but also to related subject headings. Since clicking multiple hyperlinks from one article to another is substantially easier than moving from one row of shelves to another, the user can use a single digital document as a jumping off point to a pattern of discovery that he or she would not have endured in a bricks-and-mortar environment.
I agree that the enhancements suggested in the original post above would be potentially useful, but we should not overlook the already transformative potential of digital content. While librarians and highly-trained library users may understand the progression of subject coverage in shelving by LC or Dewey classification, the average user will never comprehend those nuances. A hyperlink, on the other hand, has become in little more than a decade a widely recognized and accepted means of opening doors to additional discovery.