You are here

Unsupervised Learning of Linguistic Structure: Morphology


Unsupervised Learning of Linguistic Structure: Morphology

This course will explore the use of statistical machine learning methods as a means to work on traditional scientific questions of linguistics. We will focus on the problems of word discovery, and inference of word-internal structure (morphology), and the interaction between morphology learning and inference of grammatical categories. We will explore the conditions under which such systems can be induced in the context of unsupervised learning. The students will be introduced to basic notions of probability and information theory, as well as Kolmogorov complexity and minimum description length analysis. We will examine Python implementation of the ideas studied in the course. 

Course Status: Closed

This course is currently at capacity. Login to be added to the course's waiting list.

Course Number:


Course Session:

First two-week Session


10:30 am-12:20 pm
10:30 am-12:20 pm




No specific mathematical or programming background is assumed, but general familiarity with mathematical thinking will be a plus for attendees.