Simon Fraser University
Engaging the World

Faculty of Arts and Social Sciences

SFU linguistics professor develops new gender gap tracker

February 06, 2019
Print

Men's voices in Canadian media still outnumber women's by a ratio of more than two to one.

How do we know this? Dr. Maite Taboada, professor of linguistics at Simon Fraser University, has developed a Gender Gap Tracker to analyze how Canadian media represent women’s voices. The project was developed in partnership with Informed Opinions and with the support of SFU.

Using computational linguistics, the technology behind voice-activated assistants such as Apple’s Siri and Microsoft’s Cortana, Taboada’s Gender Gap Tracker downloads and analyzes thousands of news articles across mainstream media outlets in Canada and tracks who is mentioned and quoted, providing a very accurate gender breakdown.

Used in this manner, computational linguistics and the overlapping field of text mining have already proven their usefulness in bringing about meaningful social change.

Similar applications have analyzed police camera footage during traffic stops, showing a racial disparity: police use less respectful language with black versus white individuals, regardless of the officer’s own race or the severity of the infraction.

As a result, the Oakland Police Department updated its training modules, and other police departments in the United States are considering similar changes.

The Gender Gap Tracker works in three stages. First, it collects articles from various major Canadian news sources and stores them in a database. Then, it identifies people in the articles and determines whether they are speaking, and what their gender is. The third stage draws on the data provided and visualizes it.

The gender identification process assigns people to one of three categories: male, female, and unknown in situations where it’s not sure, such as with gender-neutral names or organizations.

Over the last few months, the Gender Gap Tracker has consistently shown an average of 74 per cent male sources versus 25 per cent female, with roughly one per cent unknown.

The goal for Taboada and her research team is to help decision-makers in mainstream media see how well they’re representing women’s voices.

The greater goal is to motivate journalists to achieve gender parity in Canadian public discourse by 2025.

Currently, the Gender Gap Tracker performs analysis for English-language media in Canada. Taboada and her team are collaborating with French computational linguists to expand into analysis of francophone media launching later this year.