The Robots Reading Vogue project launched following ProQuest’s acquisition of the Vogue Archive in 2011. Yale’s Librarian for Digital Humanities Research, Peter Leonard, was looking for already-digitized sets of texts for potentially interesting research projects when Lindsay King, an arts librarian, suggested using the Vogue Archive. Vogue has been in print continuously for over a century, and it is completely digitized and marked up as both text and images. With over 2700 covers, 400,000 pages, and six terabytes of data, the archive has become the means by which faculty and students have explored questions in fields from gender studies to computer science.
The Vogue Archive has been subject to several types of analysis and inquiry. Projects include cover averages, evaluating colormetric space, the development of fabrics, and so on. While some studies might not appear immediately important, there are many interesting topics being explored. One such study is in n-grams. The tool allows the user to input various key words and see their usage throughout the existence of the magazine. For example, researchers interested in the development of language used to describe feminine beauty can evaluate different terms to note their relevance and how that reflects broadly on perceptions of gender. Another interesting study is the look at advertisements in Vogue. Broken into various categories such as cosmetics, cars, and tobacco, advertisements have been graphed by their frequency over the span of magazine. Scholars again gain an understanding of what kinds of things society emphasized at a certain time, which companies have lasted, and created space to ask questions about why some companies or industries have succeeded and why others have not. Many such projects exist within the overall project of Robots Reading Vogue, and it has provided researchers with a unique data set to observe changes in society.
Robots Reading Vogue also uses a variety of technologies in its different projects. The n-gram research specifically uses Bookworm, the open-source version of the Google N-Gram Viewer. The primary source for the project is the Vogue Archive. It has been processed by being fully digitized and marked up to become searchable and usable in research. It ranges most widely in its presentation due to the number of different projects and ideas being explored. The research is most commonly presented using data visualization, text analysis, and timelines. Choosing how to present the data is totally dependent on the type of research.
The most pronounced scholarly primitives used in this project are annotating, sampling, and representing. Before doing any analysis on the contents of Vogue, it needed to be totally marked up and annotated to make it searchable and usable. In fact, based on the information given on the “About” page, one of the draws to project to starting this project was its availability and accessibility. By being previously marked up, researchers were drawn to the extensive archive. Sampling is also an important feature of the project. N-grams, topic modeling, and cover analyses all draw on samples from the text to help provide understanding about the broader field being researched, such as gender or design. The data is also represented in not only a clear but also an attractive way that reflects the style and contents of Vogue.
This project has many strengths. It presents many different projects that speak to different interests of researchers. However, I found it a bit overwhelming when I first looked at it. It takes a bit of browsing to get the feel of the website and understand what is going on. There could have been some clearer descriptions as well about what the different tools actually do and the value of the projects as a whole. They do not really give explicit explanations about their findings but rather allow the user to search it at their leisure. I would be more interested to read about what the creators actually learn and feel is important to share.
