Directors: Yonsoo Kim and Constantino Malagón Luque

Welcome to MMEDIS, where many interdisciplinary researchers pursue their motivated goals to create an automatic transcription of manuscripts with Artificial Intelligence. The group was founded by Yonsoo Kim, specialist in Medieval Medicine and Literatures, and Constantino Malagón Luque, specialist in Artificial Intelligence. Subsequently the MMEDIS was joined by many different experts and organizations.

Even with today’s sophisticated technology of pattern recognition automatic paleographic transcription of manuscripts is a challenging task. Transcribing and deciphering medieval manuscripts is considered complicated and problematic, especially due to particular characteristics of medieval handwriting, composition and conservation. In this sense the automatic transcription of such texts seems to be a great challenge to many researchers in the field of Artificial Intelligence. Nonetheless, with the collaboration of many specialists from different fields, we will be able to carry out successfully the current project.


Among various medieval manuscripts, the MMEDIS project will focus on medical documents of the thirteenth to fifteenth centuries. This project aims to do automatic paleographic transcription of manuscripts exclusively pertinent to Medicine, History of Medicine and Medieval Medicine. In medical and scientific fields researchers seek practical knowledge or alternative ways to cure diseases in these texts.

The main objective is thus to construct a system of automatic and assisted paleographic transcription of documents of the thirteenth to fifteenth centuries. This will be a good starting point to expand the current project and eventually develop a comprehensive program and system of paleographic transcription. One of the toughest challenges that a researcher faces in the field of paleographic transcription is the abundance of abbreviation in medieval documents. Apart from these abbreviations that can be classified as “generic and general”, one will find a great number of unusual contraction and peculiar acronym in medieval medical manuscripts. Consequently the second objective of this project is to apply this knowledge and incorporate it into the system of automatic paleographic transcription by building compiled dictionaries of generic and peculiar abbreviation, which will effectively solve the particular difficulties of transcribing abbreviated words in medieval medical manuscripts.

Moreover, this prototype of automatic paleographic recognition is instrumental to create a comprehensive dictionary of medieval medicine that will contextualize each medical term according to its unique meaning in medieval period. This project plans such an inclusive publication since none of existing pattern/character recognition software programs and manuals is helpful or useful enough to transcribe medieval medical manuscripts.

The MMEDIS project will develop new techniques and algorithms to adapt each stage of document recognition process into each specific context of transcription of medieval medical manuscripts.

Lastly, the MMEDIS project will construct a search system, recuperation and semantic categorization of medieval medical documents available in the internet. This system will also juxtapose the digitalized original documents and its paleography transcription in all different languages that it is published in the internet. In addition it will make available any information pertinent about the searching document found in various websites preselected in the form of mashups.

Funded Projects

1.- Enhancing Research in the Humanities Through an Integrated Knowledge Management System. Purdue Arts & Humanities Grant, Purdue University.