Kitchen Gadgets: Newspaper Recipe Data and AI
Loading...
Date
2024-05-30
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
International Federation of Library Associations and Institutions (IFLA)
Abstract
Recetas de las Américas [https://recetas.domains.uflib.ufl.edu/] is a bilingual web project where users can view, browse, filter, and print recipes published between 1954 and 1960 in the newspaper, Diario las Américas (Miami, FL, USA). Launched in October 2022, the project is currently undergoing an expansion to include more than 300 recipes, all from Diario las Américas and which are available through Chronicling America, a newspaper database managed by the Library of Congress in Washington, D.C., USA.
Recetas is a valuable project that calls attention to communities and contributors who have been historically underrepresented by traditional historical narratives, namely the Latin American immigrant community in Florida, particularly the Latina community, and women editors who were excluded from the newsroom and relegated to so-called “soft news” sections including “Del Hogar” where these recipes were published.
The lightweight and mostly free and open-source technical infrastructures for data manipulation and website generation in the original version of Recetas follow minimal computing principles and have the potential to help facilitate multilingual digital publishing by underrepresented and under-resourced communities on historically marginalized topics. In scaling up the project, however, we are forced to confront AI’s potential to perpetuate existing inequalities and create new avenues for exclusion and elision from the historical record.
In this paper, we discuss experimenting with AI tools to directly restructure, clean, and translate recipe text as well as to write Python code to perform these actions using traditional natural language processing resources. We will also present the findings and outputs from explorations of AI tools to generate PDF recipe cards and images of the recipes.
Description
Keywords
Subject::News media, Subject::Newspapers, Subject::Artificial intelligence, Subject::Multilingualism