Applied Natural Language Processing

Block seminar at THM (4 SWS), co-taught with Prof. Dr. Michael Guckert since WiSe 2021.

Covers the full pipeline from text preprocessing and statistical methods to modern language models: tokenization, regular expressions, TF-IDF, n-gram models with smoothing (Laplace, Kneser-Ney), word embeddings (Word2Vec, fastText), text classification (Naive Bayes, SVM), the Transformer architecture, BERT, and GPT.

Hands-on programming exercises in Python/Jupyter, final project presentations, occasional guest talks from industry.