Measuring Gender Bias in News Images

Analysing the representation of gender in news media has a long history within the fields of journalism, media and communication. Typically this can be performed by measuring how often people of each gender are mentioned within the textual content of news articles. In this paper, we adopt a different approach, classifying the faces in images of news articles into their respective gender. We present a study on 885,573 news articles gathered from the web, covering a period of four months between 19th October 2014 and 19th January 2015 from 882 news outlets. Findings show that gender bias differs by topic, with Fashion and the Arts showing the least bias. Comparisons of gender bias by outlet suggest that tabloid-style news outlets may be less genderbiased than broadsheet-style ones, supporting previous results from textual content analysis of news articles.