Harnessing technology to engage with archival sources


How can AI technologies and approaches advance scholarship on colonial history, particularly that of the Spanish empire and its American territories? In turn, how can the experimentation with text and image-rich colonial archives, which are not usually in the remit of computer science, advance AI methods?

The project is addressing these questions through three research areas that apply different AI technologies:

  1. Leverage Handwritten Text Recognition (HTR) technology to expedite the transcription of colonial Latin American handwritten and printed documents;
  2. Mine Historical Texts with Linked Open Data, Natural Language Processing, and Machine Learning to advance research in the combined use of these methods, and for the automated identification, cross-referencing, and extraction of information contained in large textual multilingual historical collections;
  3. Highlight Colonial Visual Culture with Linked Open Data and Computer Vision to develop techniques that facilitate the automated identification, extraction, search, and analysis of pictorial features in Indigenous maps and printed books.


We are using four digital collections to advance the application of AI technologies in the study of colonial Latin America. Three are under the aegis of LLILAS Benson:

  1. Digitized LLILAS Benson manuscripts, including the Relaciones Geográficas corpus;
  2. The digital archive of the Fondo Real de Cholula;
  3. The multi-institutional digitized book project, Primeros Libros de las Américas;
  4. From the National Archive of Mexico, is the UNESCO-AGN Mapilu collection.