Humanitext – a research project led by Associate Professor Naoya Iwata that seeks to revolutionize classical scholarship using AI – was selected for internationally competitive funding under the Humanities and AI Virtual Institute (HAVI) program. HAVI is administered by Schmidt Sciences, a U.S. non-profit organization dedicated to advancing science and technology for positive global impact.
The Humanitext project uses a large language model to search ancient texts for potential intertextual links such as instances of citation, paraphrasing, and critique. Subject experts then verify the generated information and transform it into data, encoding the intertextual connections so that they can be traced. The project is expected to produce a highly reliable AI system founded on expert-verified knowledge.
Humanitext was one of 19 projects to be selected among a total of 501 applications from 42 countries.
The full name of the project is “Humanitext: A Human-AI Co-Creation Platform for Verified Classical Scholarship.” The project is led by Associate Professor Naoya Iwata of the Nagoya University Center for Digital Humanities and Social Sciences (cross-appointed with the National Institute of Informatics). The other project members are Professor Ikko Tanaka of J.F. Oberlin University, Assistant Professor Jun Ogawa and Associate Professor Ikki Ohmukai of the University of Tokyo, and Specially Appointed Researcher Hiroki Uematsu of the National Institute of Informatics.
