070113 UE Workshop on Methodology - Introduction to Computational Palaeography (2024W)

This module provides an introduction to the computational analysis of medieval manuscripts and documents, including corpus building, metadata curation and feature extraction and analysis. In particular, it showcases how the use of machine learning, specifically Handwritten Text Recognition (HTR), in corpus building facilitates the assembly of large corpora without huge temporal or financial resources, taking students through the development of appropriate transcription conventions, the preparation of data and model training. The course also introduces the freeware corpus software AntConc, demonstrating different search techniques and how to export, manipulate and interpret the results in tandem with metadata about the texts. Examples will be taken from both Latin and vernacular manuscripts, so that knowledge of Latin or other medieval languages is beneficial (but not mandatory). The language of teaching will be English, but code-switching into other languages, especially German, may occur.