Using over 1 million tweets from Latin American users between 2019 and 2020, this project employs Natural Language Processing (NLP) techniques to analyze the tone and content of STEM discourse on social media. Our findings reveal the pervasive presence of gender stereotypes: girls and young women are more likely than their male counterparts to post and resonate with content promoting negative views of STEM subjects. Positive mentions of women in STEM largely originate from networks explicitly focused on gender equality, highlighting the underrepresentation of women’s participation in STEM careers. Additionally, our analysis uncovers a strong overrepresentation of male researchers and scientists in mentions of STEM discoveries, leaving women’s contributions largely invisible in the Latin American social media landscape. This research offers critical insights into how online narratives shape perceptions and belonging in STEM fields.

Community Detection Visualization

Check out the interactive visualization below:

If the visualization does not load, click here to view it directly. You can check more visualizations and the complete code on this repository