The DHRX is a cross-campus faculty research network designed to highlight innovative, digitally-focused academic work at the University of Pittsburgh. Bringing together researchers from a broad range of disciplinary backgrounds, the network is dedicated to exploring the creative use of digital technologies in humanities and social science research.
Currently, a major goal of the DHRX is to help build community. The members of the network are frequently in contact with one another and the group meets as a whole once per semester to discuss issues and challenges specific to digital making.
We are enthusiastic about the ways that digital technology can expand, reconfigure, and enrich our understanding of the relationship between academic inquiry and the social world.
For further information or to find out more about joining the network, please feel free to contact Alison Langmead, Principal Contact for the DHRX at firstname.lastname@example.org.
Upcoming PGH|DH (Pittsburgh Digital Humanities) Events (Click here for full calendar)
THATCamp Pittsburgh 2018
Starts: March 3, 2018
Ends: March 3, 2018
Location: Hunt Library, 4909 Frew St, Pittsburgh, PA 15213, USA
Jack Hessel Talk: "Grounding Images from a Digital Library in their Textual Contexts"
Starts: March 15, 2018, 4:00 pm
Ends: March 15, 2018, 5:30 pm
Description: <p>Newly digitized cultural heritage collections contain morethan just text. The availability of image data offers promising opportunitiesfor interdisciplinary work across computer vision, natural language processing,and the digital humanities. In this work, we examine a dataset of 400K imagesderived from scans of 14th-20th century volumes provided by the BritishLibrary. We combine language processing and computer vision algorithms to learna common multi-modal representation of images and the text on pages surroundingthose images. Our results suggest that processing images and text jointly,rather than independently, leads to higher quality models. While this approachmay enable scholars to search for concepts across and between textual andvisual spaces, retrieval performance on digital library datasets lags behindretrieval performance on canonical computer vision datasets. To betterunderstand the success and failure cases of this method, I will discuss recentwork that characterizes the geometric properties of image/text feature spacesin an interpretable manner.</p>
Jack Hessel Workshop: "Getting started with images in the digital humanities: What tools? Which data? And, most importantly: what questions?"
Starts: March 16, 2018, 10:00 am
Ends: March 16, 2018, 12:00 pm
Location: Digital Media Lab, 435 Cathedral of Learning, University of Pittsburgh
Description: <p>In this workshop, we will discuss the basics of how toprocess images using state-of-the-art convolutional neural networks. Thanks toopen source toolkits, this can be accomplished in just a few lines of pythoncode on your laptop. Next, we will discuss what image datasets researchers havecompiled and what questions researchers have addressed using image processingtechniques. Finally, we will discuss what sorts of new questions might bepossible to answer, with a particular focus on discussing the work of workshopattendees. Time permitting, we will also explore a handful of"first-pass" questions that one can ask about historical image datawith demos.</p>