Director, JSTOR Labs
Associate University Librarian for Research and Learning
Earlier this year, JSTOR Labs, an experimental product development group at JSTOR, released Text Analyzer, a new way to search in which users can upload their own document to initiate a search to find similar articles on the same topics. Scholars can upload near-finished manuscripts as a way to complete a literature review, and students can enter a few pages of a work-in-progress paper to find scholarship they’ll need to finish their paper. Text Analyzer uses natural language processing to figure out what the uploaded document is “about” and then recommends articles and chapters in JSTOR about the same topics. Since its release, the JSTOR Labs team has worked with Columbia University Libraries to encourage the tool’s usage and to explore possible applications of the tool. In this session, we will demonstrate the tool and the technology that powers it, share reactions of students and scholars who have used it, and reflect upon the challenges in driving adoption of a new kind of search, when users are accustomed to a single manner of interaction. We will also propose applications for this technology beyond the JSTOR corpus. These possibilities include the augment of other, current library systems, such as using a common infrastructure to create a discovery layer and aggregation of institutional repositories.