Mita Williams introduces the concept of a “Discovery Layer” in the New Jack Librarian blog and provides a reference to the work of the Discovery Tools Sub-Group of Scholars Portal, a partnership of Ontario academic libraries providing search-related products and services. Discovery layers seek to extend the functionality of federated search to improve user search experience, especially the relevance of results of the majority of queries, which are under-specified (just two or three words.)
Williams notes a number of ways in which relevance ranking can be improved:
- by taking into account the user’s previous searching behaviour
- by weighing results by the number of times an item has been bookmarked, printed, or saved
- by using citation information to determine ‘likeness’ (e.g. based on a percentage of shared citations in item’s bibliography)
- by using user-created lists articles to generate similar items of possible interest
- by knowing what courses a users is currently taking/teaching and emphasizing relevant resources accordingly
The concept is not new; use contextual information from the user’s profile, search history, plus aggregated information from members of the user’s community, to guide searching and to influence relevance of search results. As much sense as the discovery layer idea has, it has yet to gain traction in the market, although research and experimentation is underway.
The Scholars Portal Discovery Tools Sub-Group web page provides a snapshot of one major effort to improve the user search experience. The “Metasearch lay of the land” section of the page identifies federated search vendors and products the sub-group is aware of. The “Next generation discovery environment” section identifies a number of next generation catalog interfaces they are considering. And, the “User Study” section lists a number of important questions that any organization seeking to improve its user search experience would be wise to ask.
I acknowledge the tremendous effort that the Scholars Portal community is taking to evolve the ways that they support research and scholarship. The efforts of this community are worth watching as they are pioneers in creating a way of searching that will inevitably combine crawling, federated search, OPACs, and allow access to many diverse types of content, all from a Google-like interface, while providing relevance ranking that is much improved over today’s offerings.