Cal State Fullerton humanities student Michelle Ramirez is using supercomputing to preserve the centuries-old Ladino language — with the help of a chatbot named “Estreya Perez.”
Ramirez participated in the university’s Titan Supercomputing Center student research program and learned about integrating high-powered computing, data science and artificial intelligence into her project to revitalize vanishing languages.
“By stepping into high-performance computing, I hope to show that students from the humanities belong in the world of data just as much as anyone else,” said Ramirez, a double major in Spanish and liberal studies.
Ladino, also called Judeo-Spanish, is an endangered language originally spoken by Jewish people from the Iberian Peninsula who were exiled in 1492. It is estimated that about 50,000 people — mostly older generations — speak Ladino today, said Ramirez’s faculty adviser Carlos Yebra López, assistant professor of modern languages and literatures.
Yebra López, a Ladino cultural expert, helped develop the first AI-powered Ladino chatbot. He and his collaborators, computational linguist Alp Oktem and romance language expert Alejandro Acero Ayuda, launched the “Estreyika” project, which translates to “little star” in efforts to guide future generations of Ladino learners.
Yebra López said the Ladino-speaking chatbot is trained with religious and cultural references. “Estreya Perez” engages in conversation with prompts in English and Spanish, and replies to questions about daily life and community dynamics.
Ramirez wanted to be part of the project, “Revitalizing Endangered Languages With AI: Estreyika, an Interactive Ladino Teacher,” because of the impact the chatbot could make on the Ladino-speaking community.
“We hope it becomes a model on how technology and the humanities can work together to preserve endangered languages with even more limited digital footprints,” said Ramirez, whose career goal is to become a state-certified interpreter. “Most importantly, computation can be a powerful tool for preservation and accessibility.”