previous up

Seminar: Cartes Blanches (2021–23)
Intersecting Sentiment
Lukas Schwartau

A data visualisation of intersectional discrimination experiences based on a dataset of reddit posts analyzed on sentiment through ChatGPT.

The project Intersecting Sentiment aims to understand and visually represent intersectional discrimination — a concept that recognizes the interconnectedness of various forms of discrimination based on multiple social identities.

By employing data visualization and AI analysis methods, the project seeks to shed light on the complex interconnections and provide new perspectives on the multilayered nature of these discriminatory experiences. To accomplish this, an extensive dataset was curated from discrimination experiences shared on the online forum Reddit. Using the ChatGPT API, the dataset was analyzed for sentiment, examining language, emotional content, and forms of discrimination. The analysis results were then translated into a series of moving posters created with basil.js, each showcasing different layers of data visualization.

The project aims to contribute to the ongoing debate on discrimination and social injustice and encourage new insights into the intricate nature of intersectional discrimination. While this project serves as a conceptual attempt and provides a foundation for future research, there are opportunities for further investigation. To ensure the reliability of the AI sentiment analysis and consider alternative data sources, further research would be necessary.

In conclusion, Intersecting Sentiment offers an approach to understanding and representing intersectional discrimination. By combining data visualization, AI analysis, and societal interpretations, the project strives to create opportunities for dialogue, empathy, and progress towards a more inclusive society.