4 (2VL+2UE)

Basics:

Speech and Language Processing. 2022. Dan Jurafsky and James Martin, 3rd ed. draft online (https://web.stanford.edu/~jurafsky/slp3/)

The Handbook of Computational Linguistics and Natural Language Processing. 2010. Alexander Clark et al. (editors). Blackwell Publishing Ltd (https://onlinelibrary.wiley.com/doi/book/10.1002/9781444324044)

Foundations of Statistical Natural Language Processing. 1999. Chris Manning and Hinrich Schütze. MIT Press (https://nlp.stanford.edu/fsnlp/)

For more advanced literature, see lecture slides.

Skills:
Students gain an overview of the main concepts, research questions and methodological frameworks in computational linguistics. The course covers the areas of morphology, syntax, semantics and pragmatics and presents the core methods and challenges for language processing in these subfields of CL. Students learn to apply a broad range of tools in the field and discuss their benefits and limitations. Students also gain insights into a number of current topics in applied computational linguistics, such as Machine Translation, Question Answering, Chatbots & Dialogue Systems and Search.

Abilities:
Successful candidates understand the general challenges that language poses for automatic processing. Based on their knowledge across subfields of CL, they can discuss the ways in which linguistic information can be encoded for computational modeling and they can also identify those methods that are most appropriate for procesing it. For those areas of applied computational linguistics that are covered in the course, students understand the standard approaches, challenges and limitations of the state of the art.

Competencies:
Successful candidates can transfer their knowledge in computational linguistic modeling to different settings, languages and research questions. They are able to reflect on everyday computational linguistic applications like virtual assistants and machine translation systems. They can also provide a preliminary judgement as to what extent particular applications require more in-depth computational linguistic modeling.

6