Statistical natural language processing

Many core applications in modern information society such as search engines, social media, machine translation, speech processing and text mining for business intelligence apply statistical and adaptive methods. This course provides information on these methods and teaches basic skills on how they are applied on natural language data. Each topic is handled by a high level expert in the area.

Learning outcomes

After attending the course, the student

• knows how statistical and adaptive methods are used in information retrieval, machine translation, text mining, speech processing and related areas to process natural language contents.
• can apply the basic methods and techniques used for statistical natural language modeling including for instance clustering, classification, Hidden markov models and Bayesian models.

Course material

• C. Manning, H. Schütze, 1999. Foundations of Statistical Natural Language Processing. The MIT Press.
• Lecture notes

Teaching schedule

• Lectures on Tuesdays at 12–14 (10.1.–14.2.2023 and 28.2.–4.4.2023)
• Exercises on Thursdays at 14–16 (12.1.–16.2.2023, 2.–30.3.2023 and 13.4.2023)
• Exam on Tuesday 18.4.2023 at 12.00–15.00

Lectures will be held on campus. However, lecture recordings from spring 2022 and lecture slides can be provided if needed. Exercise sessions will be held on campus, but they are not mandatory.

Exercises need to be submitted around 2 weeks after the corresponding lectures. Parts of the project work need to be submitted every few weeks. The final schedule will be decided at the beginning of the course.

Completion methods

Examination and exercise work.