Researchers at the Oxford Internet Institute are building a "map of sexism" by using natural-language processing (NPL) algorithms and topic-modeling techniques to analyze data from the Everyday Sexism Project.
Founded in 2012, the Everyday Sexism Project collects anonymous descriptions of everyday sexist behavior from around the world. The project currently features websites in 10 languages and has collected more than 100,000 descriptions of sexist events.
The use of NPL algorithms and topic modeling will enable the researchers to describe patterns in the descriptions without having to read every account, or even understand the language they are reported in. Taha Yasseri, a computer scientist working on the project, says some patterns already have become clear. For example, Yasseri says he was surprised to find many of the complaints related to harassment occurring in high school, or related to classes, teachers, or schoolmates. "The striking thing was that sexism was so close to us; it's something very common," he says. "What I found fascinating was how easily this project could bring this very important issue to the public's attention."
The researchers hope the project will enable them to map out the kinds of sexual harassment and where they tend to occur. They plan to release their findings on International Women's Day on March 8, 2016.
View Full Article
Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA