Academic libraries are investing in new research products called discovery layers that strive to duplicate’s Google’s “one search box that rules them all” for the scholarly landscape. But are such efforts premature? There have not been demonstrable evidence that show the search results from such discovery layers are meaningfully relevant. This is a critical failing as the majority of our readers rarely go beyond the first page of results from any search box.
I’d like to learn from others how libraries can make better use of its data to build online services that get better the more people use it.
Libraries have been late to apply the lessons of social networks into their systems. In fact, we are still waiting for some of the most simplest applications: most library catalogues aren’t able to sort items by the number of times an item has been circulated – even when this information is available in the system.
Here’s another damning example: every semester academic libraries add books and articles into what are unfortunately known as ‘course reserve systems’ and this valuable information — that these items have been personally recommended by faculty for class use — is simply thrown away. But what if this information could be captured and shared among other academic libraries? That kind of canon could emerge over years of collection?
There are opportunities to improve this situation. With the development of open source library systems such as Evergreen, VuFind, and Blacklight – librarians are finally able to access and even adjust relevance rankings. And there is at least one discovery layer that has been designed to collect and share aggregated user information from many libraries.
I’m particularly interested if any Great Lake Campers can foresee how future developments with Zotero Commons could be integrated with library systems to make both systems more relevant to the research work of our readers.
