M.A. specialization in Computational Linguistics |
Topics
Faculty
Lab
Program Website
Website for this flyer
Introduction and Statement of Purpose
Computational Linguistics is an interdisciplinary field principally
combining elements of linguistics and computer science. Also relevant
are issues in cognitive and developmental psychology concerning
language acquisition and processing and various heuristic learning
and search techniques taken from the fields of machine learning and
artificial intelligence.
Through 5 decades of research in theoretical and applied language processing systems, the field has built up a well-defined body of results and a well-articulated set of research problems, culminating, over the last 10 years, in a period of explosive growth, both in jobs and sophistication. There are two immediate sources of this growth, first, the sucessful expansion of information technology into every corner of our lives, and second, a series of recent successes related to the advances in speech recognition technology in the 80s and 90s.
The result has been not only a growth in the number of jobs but also in the number of application areas in computational linguistics. Natural language or speech interfaces have been developed or are in development for database programs, document search engines, operating systems and software desktops, computer games and interactive computer simulations such as battlefield simulations. Natural language capability has thus become an important component of the study of human-computer interfaces, as speech and text input become an increasingly important auxiliary to graphical and multimodal systems. In addition as integrated circuit technology has migrated out of the computer and into every corner of our lives, speech and language technology has been or is being integrated into toys, household applicances, medical instruments, air traffic control systems, and GPS navigation systems.
Responding to the the University's goals of embracing technology, developing community partnerships, and heightening interdisciplinary connections, the Linguistics Department has designed a new M.A. specialization in computational linguistics. This specialization will provide a coherent array of courses leading to a level of expertise sufficient for employment in industry, thereby significantly broadening, in a direct way, our service to students and the community. The new specialization is designed also to provide a strong foundation for Ph.D study.
Faculty
Jean Mark Gawron. Jean Mark Gawron has been at a number of centers of computational linguistic research, including University of Edinburgh Epistemics and AI programs, the New York University Courant Institute, the Center for the Study of Language and Information at Stanford. For a number of years he was a member of the Natural Language group at SRI International. He joined the Department of Linguistics and Oriental Languages to start up a computational linguistics program in the fall of 2000. His research interests include semantics and pragmatics, probabilistic parsing, interleaving of syntactic and semantic constraints, Machine Translation, and computational lexical semantics.
Web-site: http://www.rohan.sdsu.edu/~gawron.
Rob Malouf Rob Malouf is currently a post-doc in the humanities computing department at the University of Groningen in the Netherlands and will be joining the SDSU Computational Linguistics program in the fall of 2002. Rob was a member of the NWO PIONIER research project Algorithms for Linguistic Processing. Before that, he was involved in the project "Computational and theoretical modeling of structures in performance," part of the School of Behavioral and Cognitive Neurosciences at the University of Groningen. His work is aimed at integrating symbolic grammars with knowledge derived automatically from corpora via machine learning techniques. He is particularly interested in data-driven approaches to modeling the development and behavior of syntactic categories. Recently, he did some work on using memory-based learning to extract rules for ordering prenominal adjectives in English from large corpora. His current work focuses on applying maximum entropy/minimum divergence modeling to attribute-value grammar formalisms such as HPSG using big computers. He is also interested in other applications of probability theory.
Web-site: http://odur.let.rug.nl/~malouf/
Courses
- Ling 570: Mathematical Linguistics (new course), or equivalent
- Ling 571. Computational corpus linguistics (new course)
- Ling 581. Introduction to computational linguistics (new course)
- Ling 582. Computational syntax and semantics (new course)
- Ling 596. Selected topics in linguistics (with CL content)
- Ling 620. Advanced formal syntax (new course)
- Ling 654. Language and cognition
- Ling 681. Statistical methods in computational linguistics (new course)
- Ling 682. Computational sentence and discourse semantics (new course)
- Ling 696. Selected topics in linguistics (with CL content)
- Ling 795: Seminar in Linguistics (with CL)
Computational Linguistics Lab
A key component
of the computational linguistics program at SDSU is the integration
of practical lab component in most courses.
The computational linguistics lab, located in the Professional
Studies and Fine Arts Building, is housed as a
separate lab within the Social Science Research Lab.
The lab
consists of a subnetwork of 15 Pentium III PCs running
Red Hat Linux with RAMs ranging a Half a Gig to 2 Gig.
There is also a Sun Ultra 80. The PCs are equipped with
read-write CDs, microphones, sound and video cards,
and individual hard disks.
Over and above the actual hardware the language modeling lab has three
components.
For more detailed information about the lab, visit the lab
website:
http://bulba.sdsu.edu/research/description.html.
Website for this flyer