Robots Reading Vogue is a website that data-mined thousands of Vogue magazines, allowing users to interact with the data and learn about different topics. Yale’s new Librarian for Digital Humanities Research, Peter Leonard, wanted to use a dataset that could encourage new digital humanities projects, leading to the creation of the project. On this website, users can search various things like the “shifting notion of beauty” through word vectors, or search and compare the usage of words across Vogue’s entire magazine run.
The source of this media is the Vogue archive (created by Proquest and Condé Nast in 2011), which collected over 2,700 covers, and 400,000 pages, in total being around 6TB of data!
One of many cool ways the dataset was used was Vogue Covers in Colormetric Space, which used a free tool called ImagePlot, to explore the visuals of Vogue covers through their hue, saturation, and brightness. From the 1890s to 2010s (left to right), we can look at Vogue’s visuals over time, with the more colorful visuals being on the higher y-axis.
The image below provides a better look at the interactive graph! The actual graph is extremely big (about 20,000 pixels wide), so this only covers only about two decades of what’s actually available.
I can definitely see this being used in an assignment about fashion trends and how they contribute to the gender identities of today. I’m not entirely sure where to find the code for this project, and because of a software design class I took this time last year, I do have a solid grasp of how they manipulated the data on the backend, but I am very confused about how it translated to the visuals on the website.