Professor and Associate Dean for Research J. Stephen Downie will present his recent work with the HathiTrust Research Center (HTRC) on April 30 at Harvard Library. Downie is codirector of HTRC, a collaboration between the University of Illinois, Indiana University, and the HathiTrust to enable advanced computational access to text found in the HathiTrust (HT) Digital Library.
His talk, "Creating Universal Open Access to Closed Textual Data at Scale: Use Cases from the HathiTrust Research Center," will discuss how the HTRC is creating a set of non-consumptive research services to make HT Digital Library volumes that are under copyright restrictions more open and useful to scholars.
"The creation and publication of the HTRC 'Extracted Features' (EF) dataset provides unigram counts and Part-of-Speech (POS) information for each of the 5.6 billion pages in the HT Digital Library," explained Downie. "In my talk, I will introduce two uses cases that leverage the EF dataset: the 'HathiTrust + Bookworm' visualization and analysis tool; and the Workset Building environment developed to provide researchers fine-grained access to the entire HT collection (both public domain and in-copyright) via the EF dataset."
Downie leads the HathiTrust + Bookworm text analysis project, which is creating tools to visualize the evolution of term usage over time. He also is the principal investigator on the Workset Creation for Scholarly Analysis + Data Capsules project, which integrates workset models and tools, and he represents the HTRC on the Novel(TM) text mining project as well as the Single Interface for Music Score Searching and Analysis project. All of these projects strive to provide large-scale analytic access to copyright-restricted cultural data.