Historical Markup Tool

The historical markup tool was developed in 2019 by Audrey Holmes as part of her Master's Thesis in Computational Linguistics at the University of Washington. The tool accepts raw text as input and generates TEI output with named entity labels and references for disambiguation.

The tool was trained primarily on 20th century English documents but should generalize to English documents from most time periods. The Flair Python library is used for named entity recognition, along with domain adapted word embeddings for entity resolution.