The Applied Computational Linguistics group at Potsdam University conducts research on different perspectives on discourse structure (coherence, coreference, etc.) and on applications of discourse processing, for example in argument mining and in collaborations with social sciences (educational, political science). See http://angcl.ling.uni-potsdam.de.
The position to be filled is in a new DFG-funded project with partners in Computer Science (Halle-Wittenberg) as well as Literary Science and E-Research (Berlin). The goal is to implement a solution for text alignment in scenarios where an author writes different versions of the "same" text in different languages; i.e., where portions of the text in the other language can be changed, shortened, extended, or reordered. In collaboration with the team in Halle-Wittenberg, the successful candidate will conduct experiments on building alignments on word, phrase and sentence level. A central basis for these alignments will be measures of semantic similarity, for example by utilizing multilingual word embeddings.
The position runs for three years and offers the possibility for obtaining a PhD. Salary is according to the Brandenburg TV-L salary scale (75%), see for instance https://zbb.brandenburg.de/sixcms/media.php/9/Tabelle%20TV-L_2022.pdf
Candidates are expected to hold an MSc degree in Computational Linguistics or a related field, with experience in methods relevant to the project. Solid programming skills in Python are required, experience in working with (multilingual) corpora is a plus. Near-native knowledge of either English or German is required, and some knowledge of the other language is a plus.
Please send your application before Nov 17th to the email below, consisting of a CV, a statement of research interests and a summary (or a sample chapter of) the MSc thesis; letters of reference may optionally be added.
Inquiries and applications to: Prof. Manfred Stede, stede@uni-potsdam.de