Possible Research Questions

The main goal of this research is to provide codicologists better insights into high-dimensional datasets of digitized medieval manuscripts by visual analysis. Currently, only standard manual techniques exist to analyze medieval manuscript pages. For example can be recorded manually in which positions throughout a codex the number of pictorial areas increases compared to text area. Otherwise can be recorded, if the size of text space and margins is changing throughout the codex. In this manual approach, the quality of the analysis depends on the amount of the data that can be measured manually, which means that more observations of numeric and categorical data can lead to better analysis and large data tables. However, these large number of data tables provide more time for the researcher during the analysis stage. Thus, better visual analysis is required to make the subjective analysis of codicologists objective.

A statistical evaluation allows the empirical analysis of the manuscript pages. By building correlations between metadata from the manuscript descriptions (e.g. dating, text genre, writing material) and metadata from the automatically acquired layout information new research questions can be addressed to the collection, such as:

•  detection of dependencies  and correlations between written space/white space and codex format

•   evaluation of proportions between pictorial and written space

•   tracing of references between content and page layout

•   joining of codex fragments on the basis of the page layout

•   identification of writers by page layout