136038 UE Doing research with text corpora (2024S)

The course introduces students to the study of text corpora. A corpus, in its broadest sense, is a structured collection of texts. In modern usage, this usually refers to a digital text collection that is annotated with respect to a pre-defined set of analytically relevant features. Although the systematic study of machine-readable text corpora as an empirically based method has mostly been developed within the field of linguistics, text corpora can be useful for investigating all sorts of research questions within the Digital Humanities.
The course will first introduce students to a number of corpora that are available online. Students will learn to apply various browser-based search and analysis tools. Next, students will learn how to compile and annotate their own corpus using tools based in machine-learning. Students will become acquainted with the various formats that different corpora are encoded in, with a particular focus on XML formats. Students will learn how to apply various methods for analysing corpus-derived data, including basic statistical testing, regression modelling, and network analysis. They will also learn how to visualize their results and present their research in the form of a poster or oral presentation.

The software used for all corpus construction and analysis is R (including wrappers of Python-based tools).

The approach is both theoretical and practical, with hands-on exercises in project planning and prototyping. Students are expected to have some familiarity with digital environments, and previous practice with programming is desired, but not strictly mandatory. The course will be mostly held in English, but some code-switching between English and German will invariably occur.