The main objective of this course is to promote awareness and 
familiarity with the current state of the art in computational 
linguistics, which is a challenge in a field where hundreds of 
new research results are published every quarter. To meet this 
challenge, students will read and discuss papers in two 
categories:

a. foundational – journal-article length, reviews an entire 
subfield

b. research – conference-paper length, reports a cluster of 
related results

Building on the knowledge they have acquired in earlier courses, 
such as A Finite-State Introduction to Computational Linguistics 
and Natural Language Processing Techniques, students will 
approach the research frontier step-by-step as a cooperative 
group. They will learn to critically analyze current research in 
the field, synthesize this material, and conduct their own 
investigations. Each participant will complete a final project 
that includes a literature review. In this way, the seminar will 
help prepare students to carry out thesis or dissertation 
research.

The specific choice of topics will vary from year to year and 
may include: 

- incremental parsing in various grammar formalisms
- semantic role labelling
- vector space models of semantics
- machine reading, question-answering, and successors to 
information extraction
- machine translation and multilinguality

All papers will be available electronically either via the UGA 
library or from the ACL Anthology 
https://www.aclweb.org/anthology/.