New corpus standardizes medieval French medical texts

Researchers have developed ALMeD, a semantically rich corpus of medical and surgical treatises written in Old and Middle French.

This new dataset provides a gold standard for analyzing medieval medical literature, which has previously been overlooked in favor of legal or literary texts. The corpus includes texts written in Medieval French, a Romance language that heavily influenced other vernaculars of its era. Unlike previous natural language processing efforts that relied on non-scientific sources, ALMeD focuses on specialized medical terminology.

The data accounts for the complex morphological development, changing syntactical structures, and inconsistent spelling common in these historical languages. The project aims to improve how computers parse and understand historical scientific documents. While the corpus is currently a work in progress, it offers a more precise foundation for linguistic analysis than existing resources. These tools help scholars better understand how medical knowledge was recorded and transmitted throughout history.