Joe Loewenstein, a professor of English and director of the Humanities Digital Workshop, both in Arts & Sciences, will serve as co-principal investigator for a $798,000 Human Networks and Data Science grant from the National Science Foundation. Seth Kulick and Neville Ryant, both of the Linguistic Data Consortium at the University of Pennsylvania, will serve as co-PIs.
The project, entitled “TRACE: Enriching EarlyPrint,” builds on the EarlyPrint Library, a digital collection of 60,000 books published in English between 1473 and the early 1700s, which Washington University in St. Louis curates in partnership with Northwestern University. The grant will enable researchers manually to annotate samples from across the EarlyPrint corpus, comprising up to 1 million words; to train new natural-language processing models on those annotated samples; and then to deploy those models automatically to annotate the entire EarlyPrint corpus of 1.65 billion words. Once the resulting annotations have been assessed, the group will reconstruct the EarlyPrint Lab, an associated site that offers a range of computational, analytical and visualization tools, for the scholarly investigation of this immense library.
The new annotation layers will encompass, among other topics, word-sense disambiguation; syntactic analysis; and co-referencing chains of text to corresponding events and such entities as people, places and organizations. Once complete, the annotated corpus will effectively provide a unified index of all the people and events that imprinted themselves on 16th- and 17th-century English-speaking public discourse.
The WashU team will be housed within Arts & Sciences’ Humanities Digital Workshop. Supervised by Loewenstein’s colleagues Douglas Knox and Tumaini Ussiri, annotators will include students in English, history, linguistics, statistics and data science. Serving as a consultant on the grant will be WashU alumnus John Ladd (PhD ’19). Now an assistant professor of computing and information studies at Washington and Jefferson College, Ladd has been working on EarlyPrint since 2012, when he entered WashU’s doctoral program in English.
Read more about the project on the Humanities Digital Workshop website.
