Prerequisites: (LING 1069 OR LING 1200) AND (LING 3300 OR CS 1410).
An introduction to computational linguistics for students with previous programming experience. This course explores the models, algorithms, and techniques that dominate modern-day language technology, and it evaluates them from a linguistically informed perspective. Topics include corpus-based methods, finite-state approaches, word vectors, computational models of cognitive processes, and model evaluation techniques. Great emphasis is put on discussing the limitations of existing techniques and how they might benefit from linguistic insights. We also discuss Ethics in AI and in Natural Language Processing, and the way fairness considerations should inform data collection and algorithm design. Students will also hone their programming skills and develop familiarity with state-of-the-art software packages for computational linguistics.