Let’s talk about debiasing books – Webinar

Let’s talk about debiasing books – Webinar

As part of BookNet Canada‘s on-going webinar series, I recently held a webinar on using data analytics to “debias” books. A lot of discussion has begun in publishing with respect to increasing the diversity of publishers’ lists. The issue is often seen as one that…

Are you genre fluid? A new collaboration on Spotify by Cheng Lin and Benjamin LeBrun

Are you genre fluid? A new collaboration on Spotify by Cheng Lin and Benjamin LeBrun

A new student collaboration is out. It’s called “Streaming Bias: studying music curation on Spotify” and represents the lab’s first attempt at studying online music content. Lin and LeBrun did an amazing job in conceptualizing the project and working tirelessly from data collection to analysis…

Are academics more unequal than athletics? A new collaboration by Brian Powell

Are academics more unequal than athletics? A new collaboration by Brian Powell

Our latest lab collaboration was created by McGill undergraduate Brian Powell. As a college basketball fan, he was interested in whether the high levels of institutional inequality within academic publishing and hiring that we have been seeing would also hold true within college athletics. Did…

Let’s talk about debiasing books

Let’s talk about debiasing books

This March I will be giving a talk at BookNet Canada’s TechForum. The topic is “debiasing” books and I’ve posted a short blog post to get the conversation started. Discussions of bias, and cultural bias in particular, have been much in the air lately. A…

Social Characters: new lab collaboration by Eve Kraicer

Social Characters: new lab collaboration by Eve Kraicer

Eve Kraicer’s lab collaboration (our name for working papers) is now out! It’s about the ways in which gender is represented in contemporary fiction looking at a collection of ca. 26,000 characters. Some of the salient findings: the ratio of women to men characters in…

How do we Model Stereotypes without Stereotyping (Again)? How about information theory.

How do we Model Stereotypes without Stereotyping (Again)? How about information theory.

In a previous post, we explored how using language models and the idea of “perplexity” can allow us to study stereotypes in movie character roles using their dialogue as a basis. We examined a corpus of 750 Hollywood films, released between 1970 and 2014, and…

Does the decline of gender within literary studies matter?

Does the decline of gender within literary studies matter?

A little while back I posted on the unmoving ratio between male/female pronouns in a data set of ~60,000 articles in literary studies. The ratio has been stable at about 2:1 since the mid-1990s. While pronouns do not tell us everything we need to know…

How do we model stereotypes without stereotyping?

How do we model stereotypes without stereotyping?

We recently put out a paper on how racial bias functions in Hollywood films. This work was based on a few studies that came before it, namely this one, from USC Annenberg. We presented numerical analyses like the number of characters in different racial and…