If You Give a Robot the Vogue…

It’ll probably learn to read it.

The project I had explored was titled “Robots Reading Vogue,” conducted at Yale University in 2014. From topic modeling to color analysis, the whole project contains many different components that utilize the given data in various ways. I was specifically interested in the n-gram search, which was accessible from a grant given by the library’s Standing Committee on Professional Awareness (SCOPA). The proposal, “Implementing N-Gram Search Across Yale Digital Collections” was approved and applied to the Vogue archives.

Source

Vogue Archives

Processed

Bookworm: the open-source version of the Google n-gram viewer bookworm.culturomics.org

Presented

Visual N-Gram Graph

Who made the website? What are their relationships to the institution?

The librarian for Digital Humanities Research at Yale, Peter Leonard, was the spearhead for the project. He was searching for a large set of digitized data, both textual and visual, to spark an interesting digital humanities project. Lindsay King, an art librarian, helped him to start the project.

What kinds of data are being used? Is the data available for broader use? Would you want it to be?

The project used 6TB of data from Vogue archives to find statistics. They looked at cover pages for the magazine, textual pieces, and visual images to complete the work. It was made available to the Yale University Library by a perpetual access license, which was purchased in 2011. Because you need to have a license for access to the data, it is not available to the majority of the population. I would not find it necessarily useful to the broader public unless an individual was well equipped in knowledge as well as resources to handle the copious amounts of data.  

What are the components?

The n-gram search visually explored lexical trends in the magazine through the years, creating statistics for the number of times certain words appeared in the text. It is represented as a line graph.

Which academic fields do you see the project in conversation with?

From the data collected from the n-gram search, I can see the project in conversation with both cultural and gender studies. The words analyzed in the text were “lovely”, “pretty”, “beautiful”, and “sexy”. In the statistics, we can see a trend of what words were used during different time periods. This can lend itself to both cultural and gender studies in the sense that we can understand what words were commonly used to describe women at the time and how cultural shifts change that. For example, the word “sexy” was not used to describe women until the end of the 1960s, where it began to increase substantially. This could be developed by the idea of a more conservative past (sexy is scandalous) where feminine beauty was viewed differently than its more modern and liberalized counterpart.