128141 FS FS Research Seminar I / II (2021S)

Authorship attribution is concerned with identifying the author of an – often anonymously written – text. The term subsumes a variety of methods, both qualitative and quantitative, that are used for diverse applications in many different fields such as historical linguistics (e.g. identifying the author of a historical text), forensic linguistics (e.g. identifying the author of fraud mail), or plagiarism detection. In this course, we will focus on quantitative methods from the field of stylometry. Students will learn how to employ statistical programming, methods from text mining and basic machine-learning techniques in authorship attribution tasks. Students will work in groups on small projects, in which they first compile their own training corpus and then apply relevant methods to tackle anonymous texts.

This course has a quantitative focus. Statistical programming will be done in R(Studio Cloud). In this regard, no preliminary knowledge is required (introductory tutorials will be provided), however, previous exposure to R (e.g. in the MA English Research Methods course) will prove useful. I will take knowledge of high-school mathematics (basic calculus, percentages/fractions, probabilities, functions like log or exp) as well as of linguistic concepts (word classes, types/tokens, etc.) for granted.

All classes will take place online. There will be individual feedback sessions for each group (dates to be announced).

IMPORTANT: All projects will be presented at a student-mini-conference, which will take place online on Saturday, June 19th, 10:00 to 14:00! Attendance is required.