Discover the amazing data work McGill students are doing! This panel will feature: 

Sil Hamilton - Parallel Worlds: How To Generate Useful Data with AI

Creative AI models like ChatGPT or Dall-E are becoming increasingly capable, but ways to integrate them into your empirical research process remain unclear. Discover a methodology for those in the humanities and the social sciences: generating synthetic datasets for comparative purposes. This session will report on a published project in which we use AI to evaluate editorial framing in the news media, and reflect on how AI made our approach possible.

Heather Rogers - Digitizing Botanical Herstory: The Botanical Collections of Dr. Dorothy Swales

This project features the botanical collections of Dr. Dorothy Swales, the first woman to curate the Macdonald College (later McGill University) Herbarium. Dr. Swales’ life was inseparable from the natural world and telling her story as a trailblazing woman in botanical curation necessitates weaving in the lives of the mosses, lichens, and flowering plants that she dedicated her career to better understanding and preserving. The technological aspect of my research incorporates digitizing botanical vouchers from Dr. Swales’ tenure as curator, culminating in a digital exhibit using Omeka. An overarching objective of this project is to contribute to the growing field of Digital Environmental Humanities by demonstrating how digital tools can tell nuanced, interwoven stories about human-plant entanglements.

Alexander Springer - A GIS Approach to a History of Epidemics in 19th Century India

An exploration of the digitised medical reports on disease in India and the significant challenges in finding data, interpreting it and then presenting it on GIS and in spreadsheets.

Jonathan Love - NLP Scene and Short Story Collection Detection

Can we build an AI tool that can automatically detect if an input text is a singular story or a collection of short stories? This presentation will showshow how by transforming character counts into signals and then using signal processing algorithms, one can naturally divide up an input text into segments. Then, all that is left is to build a classifier and train it based on how each text was divided up.