Library Technology has been on a wild ride since the 1960s, when computer databases first became available. Automation has made so many wonderful advances possible, but there are still so many frustrations and bottlenecks that leave our users overwhelmed and stressed. While I try to discuss PKM/productivity tools more specifically, most of the time, this blog does focus on issues relating to the productivity of academic researchers. Researchers do… well, research, and the usability of library resources is, in my view, integral to their productivity.
Take search engines, for example. I did discuss this some in my last post, but the point now is the fragmentation. There are so many databases with information from so many publishers. Even a simple thing like trying to find the download pdf button for an article after identifying it in a database is frustrating, because every publisher puts it in a different place and some of them do not stand out well on the page. We all know the saga of the attempts to make searching library databases simpler and more Google-like, from federated search to web-scale discovery services (WSDS). The more I use a WSDS, the more problems I find with it. There are so many results that good items can get lost. For example, our WSDS includes CQ Researcher results. CQ Researcher is a database consisting of deep background reports on public policy topics, and the reports are rather like encyclopedia articles on steroids. So they are an excellent place for a student to start researching a topic, as what they need at the beginning of researching a paper is a mental map, so to speak, of the concepts involved. Yet CQ Researcher reports, which should be the first thing a student looks at, get lost in a search in WSDS. Relevancy ranking is better than it used to be but still not as good as it needs to be.
My main focus for this article, however, is the serendipity factor. My faculty are primarily social scientists, and we know that those in the social sciences and humanities tend more often to find useful material while browsing the book shelves in a library. More and more libraries are moving print book collections off-site, or are planning to do so. Moreover libraries are buying more electronic books and moving away from print ones. Given these factors, is it possible to create a user-friendly browsing feature?
My thoughts here are heavily influenced by the future loss of physical books, but also a few blog posts, which, of course, came into my awareness through serendipity . One is Planning for Serendipity, from the Digital Public Library of America (DLPA). It discusses the problem, links to some other useful articles, and mentions features that might possibly help serendipitous discovery in online environments. It has to be admitted that a library catalog record simply doesn’t provide enough information for most users to decide if this particular book is going to be useful for them. It would really be helpful if users could count on seeing things like the table of contents, the publisher’s blurb for the book, reviews by editorial writers and other users, and an idea of which books on a topic have been used most or cited most.
Our library uses the VuFind overlay on top of Voyager data, and it links to the Google Books preview (if there is one) and the Amazon record with its review data (if there are any). That helps, if the users (a) see it; and (b) know what it can do for them. Such things ought to be standard and in a unified interface.
Pieces of a solution to the problem exist now. Google had an interesting experiment called WebGL Bookcase. When you open it, you choose a subject and see a carousel of books:
You can then choose a book and see its cover:
Since the carousel has not been updated since 2011, Google may have abandoned it. It seems like a good start, but could use still more metadata – a link to the table of contents, at the least.
Another interesting experiment along these lines is the library catalog for North Carolina State University (NCSU). The Shelf Browse feature was put together by a couple of graduate students there. So in the record for a book, there is a Browse the Shelf link prominent on the right side:
Anecdotally I have heard recently of a professor who was quite worried about not being able to browse the books (once the library converted to electronic only) and, after seeing the NCSU browse shelf feature, thought this was something he could work with.
Finally, another interesting idea is Stack View, from Harvard Library Innovation Lab. There is a video on the page. Basically Stack View shows you a shelf of books, rotates it so you see them stacked on top of each other, and the “spine” of the book is proportional to its size (page length and length and width). Plus the books that are circulated the most are the darkest blue in color, giving a sense of the book’s popularity.
So there are a lot of ideas out there. One of the blog posts on this topic, Enabling the Research ‘Flow’ and Serendity in Today’s Digital Library Environment, discussed some the author’s ideas of what a digital library needs to offer:
If we set aside the psychological factors that contribute to the occurrence of a serendipitous discovery, what is essential to efficient browsing boils down to how easily (i) we can scan through many different books (or information units such as a report or an article) quickly and effectively and (ii) zoom in/out and switch between the macro level (subjects, data types, databases, journals, etc) and the micro level (individual books, articles, photographs, etc. and their content).
My library has just taken a great step forward in improving the usability of reading scholarly articles by subscribing to Browzine. It is an app for smartphones and tablets (iOS and Android, so far) that graphically displays journals on a bookshelf. Users select their school, then sign in with their school’s authentication and see the scholarly journals the campus library subscribes to. They can add journals to their shelf, see what issues are available, and get notifications when new issues are available. One can view PDFs in Browzine and store them for later use:
So usabililty is getting better all the time, but let’s not kid ourselves – there is still a long way to go. Ideally there could be something like Browzine that works with all of a library’s resources, and has a standard set of rich metadata so that users can quickly decide on an item’s utility. A lot of attention has been to the discovery of library resources, but less has been paid to their usability once discovered.