STC^2: scalable text collation for the short title catalogue of early printed books [STC2: scalable text collation for the short title catalogue of early printed books is a level III project to…

The National Endowment for the Humanities (NEH) Office of Digital Humanities awarded Northeastern University a $338,357 Project Grant titled "STC^2: Scalable Text Collation for the Short Title Catalogue of Early Printed Books". This Level III digital humanities project aims to build open data and code infrastructure for metadata, page images, and manual and automatic transcriptions of over 242,000 early modern printed books. The project will collate manual transcripts from the Text Creation Partnership with optical character recognition output to support more accurate OCR, text normalization, restoration of textual gaps, detection of manuscript annotations, and contextual embeddings for search and bibliographic analysis. All data and code will be published under an open-source license. This project complements the awardee's work on OCR, book search, and corpora beyond the English Short Title Catalogue. The award period is from February 1, 2025 to January 31, 2027.

STC^2: scalable text collation for the short title catalogue of early printed books [STC2: scalable text collation for the short title catalogue of early printed books is a level III project to build open data and code infrastructure for metadata, page images, and manual and automatic transcriptions of early modern printed books. We will collate manual transcripts from the text creation partnership with optical character recognition output from the microfilms of the short title catalogue of early english printed books recently scanned by the internet archive and other open-access images. This dataset will support more accurate OCR, text normalization, restoring textual gaps with evidence from comparable copies and editions, detecting manuscript annotations, and contextual embeddings of words and pages to support search and bibliographic analysis. We will publish this data for over 242,000 early modern printed books and all code under an open-source license. This project complements our team's work on OCR, book search, and corpora beyond the english STC.]