Robots Reading Vogue: How text and data mining (TDM) cast new light on a large historical archive
The ever-increasing amount of digitized cultural heritage material online has researchers and students alike curious about new approaches to making sense of large-scale digital archives. Text and data mining is one approach to meeting this challenge -- either by allowing researchers to search across vast amounts of data in order to answer specific questions, or by surfacing latent patterns within Big Data.
In this webinar, Peter Leonard and Lindsay King from Yale University Library explain how text and data mining techniques cast new light on a large historical archive: every page of every issue of American Vogue magazine from 1892 to 2013. They will start with simple techniques, such as n-gram visualizations, and move on to more sophisticated approaches such as topic modeling and word embedding. Finally, they will consider some emerging ways of applying data mining techniques to photography and illustrations.
Prior to Peter and Lindsay’s presentation, key concepts of TDM will be presented by Evangelos Theodoridis (PhD), along with indicative use cases and applications. Finally, various challenges in this area will be highlighted.