Skip to content
Stories

Seeing Our Neighborhoods: Providing Public Access to the Boston Globe Photograph Collection

Partially supported by a NULab Seedling Grant and an American Rescue Plan award.

“Photo morgues” are archival collections of photographs from a newspaper’s past issues and news cycles. Stored, or else in the jargon “sent to the morgue”, these files often remain hidden in the history of journalism, in the form of defunct and forgotten newspaper clippings. While photo morgues can support a wide range of audiences and pursuits—including academics, journalists, historians, citizens and local communities—identifying and accessing photographs of interest is often difficult for several reasons. Many physical photo morgues are maintained within the organizational scheme created by the source newspaper, requiring to search through any number of thematic folders. While digitization offers the hope of solving access issues, many digitized photo collections have been monetized, reducing the access for many potential users.

“Seeing Our Neighborhoods: Providing Public Access to the Boston Globe Photograph Collection” seeks to bring one of these collections back to life, by making openly accessible an extensive archive of digitized photographs from one of most relevant newspapers in US journalism, the Boston Globe. From local events to nation-wide social movements, the Globe’s archive of photographs stretches back for over a century and constitutes a diverse cultural record. With the support of a group of researchers, the project is committing to digitize a section of the full archive—selecting the photographic prints that help document the history of Boston, covering education, architecture, infrastructure, neighborhoods, politics, protests and demonstrations, and other major topics of interest.

With the aim to benefit multiple audiences, including students, scholars, and the general public, the project will use computational techniques to reduce barriers to the Boston Globe photo-archive, by training machine learning models for handwriting transcription and image analysis to enrich records with metadata. The goal is to create an information retrieval system able to find similar items in the collection, track trends over time, and retrace the many narratives contained in the history of Boston journalism.

Principal Investigators

David Smith, associate professor in the Khoury College of Computer Sciences; Giulia Taurino, Postdoctoral Research Fellow

More Stories

Renewable Messaging of Fossil Fuel Companies: Corporate Communication Strategies in Authoritarian Regimes

01.11.2024

Mapping Real-World Migration Flow Network Using Context-Rich Social Media Data

01.11.2024
Sample Bipartite Network of Advocates (colored circles) and cases (grey squares)

The Network of Hidden Plaintiffs in U.S. Supreme Court Litigation

02.20.24
Research Projects