Users find the the list of resources and terms for finding information bewildering: book, journal, article, database, resource, catalog, worldcat, guide, website, google scholar, institutional repository, digital collections, archives, etc. What they really want is a single search box which will return a single, relevance-ranked result set across all Libraries' resources and more. In the absence of a single data source to support such a search, Libraries for many years have tried to dynamically aggregate and de-duplicate federated searches across multiple data sources, called metasearch, which has not worked very well. In recent years a new model, often called Bento Box, has become popular which attempts to come closer to the ideal search. In this model the user enters their search into a single box, then multiple sources are searched and presented back on a single result screen, boxed into separate areas without de-duplication, with only a few results from each source. Then the user can clearly see that results have come from multiple sources and either select a specific hit or see more results from any of the sources.

Poster presentation at the University of Maryland Libraries Research & Innovative Practice Forum on June 8, 2016. The poster proposes that the UMD Libraries should evaluate adoption of Bento Box Discovery for improved user search experience.