This is the course page for the seminar course Computational Morphology at the Department of Linguistics, University of Tübingen.
This is a seminar course covering a wide range of different approaches for computational processing of natural language morphology. We will cover both rule-based (finite-state) systems and data-driven models of both supervised and unsupervised morphological processing.
The course is organized as two parts. First part constitutes a series of lectures and tutorials, where fundamental topics and tools will be introduced. The second part follows a seminar format, where we will read a selection of influential and/or recent papers from the literature, and discuss them in the class. The participants will also be required to complete small projects.
The questions of interest for this course include:
- Can we (precisely and computationally) describe morphology of a language?
- Can we, or to what extent, segment words in natural languages to segments that correspond to morphemes?
- What is a word anyway? How can we discover words in speech, or texts written without clear hints to word boundaries?
- Can we build models of morphological processing purely data-driven manner? Do we still need rule-based models/descriptions of natural languages?
- Can we help document low-resource, endangered language with the help of computational methods?
- What makes a language morphologically complex or rich? Can we quantify morphological complexity? Is it relevant to natural language processing (e.g., with modern NLP systems)?
Grading will be based on class participation and homeworks (6ECTS). Further course credits can be obtained by a term project described in a term paper (9ECTS) within the scope of the course.
Requirements
There are no formal requirements. However, this is an advanced course where the participants are expected to read and understand a broad range of topics including finite-state automata and various machine learning methods.
Contact
- Instructor: Çağrı Çöltekin
<ccoltekin@sfs.uni-tuebingen.de>
Office hours: Monday 10:00 - 11:00 (Wilhelmstr. 19, room 1.09)