This page attempts to answer some common questions we get. Email us if you have others.
An overview of aspects of our computational linguistics program can also be found in Jason and Katrin's TEACH-CL08 paper.
We also recommend looking at Mark Johnson's excellent advice to graduate school applicants interested in computational linguistics.
What are good ways of knowing whether I should study computational linguistics?
If you think language is fascinating, but you want to approach it in a formal manner that allows you to actually build interesting applications and/or rigorously test hypotheses about it via computational implementations.
Or perhaps you are coming from a strong background in algorithms or machine learning and are looking for a challenging and interesting set of problems to apply them too. Language gets tricky, fast. :)
See these pages:
I want to apply computational methods to studying language. Is what I want to do computational linguistics or corpus linguistics?
The two fields look completely different to people who are in them, but are confused surprisingly often. Here are some pages that try to define the two fields:
Would you recommend getting a graduate degree from Computer Science or from Linguistics if CL is the main thing I want to study?
A CS degree probably leaves you with more options, but having the Linguistics degree is no obstacle. A number of our Linguistics graduates have gone on to industry positions. One big difference is that CS graduate probably has a better shot at a tenure track job in a linguistics department than a Linguistics grad trying for a CS position.
As important is thinking about who you would like to work with and applying to the relevant department.
Another point is that some people discover computational linguistics late in their undergraduate degree in linguistics or a related field but haven't taken much computer science; they would likely stand a better chance of admission to a linguistics department than to a CS department. If you are a quick learner and have an aptitude for programming and math, then you can still possibly get in to the linguistics program and end up doing work that is every bit as “computational” as those who go straight into CS.
Is there anything I should be doing between now and when I apply to help my chances of being accepted?
If you still can do so, take courses that are relevant to computational linguistics.
Try to have a clear statement of purpose that discusses what you are interested in working on, as specifically as you can be. Talk about specifics of work you might have already done (including work in industry) and why it is interesting and whether you did anything particularly innovative to deal with the task. If possible, provide a good writing sample (preferably one that relates to a topic in Linguistics or CS that shows your ability to think independently and write clearly). If you are in a position to do so, make your honors thesis about a topic related to computational linguistics, or if you aren't doing an honors thesis, try to subvert a class paper so that it is as relevant as possible.
Make sure you have strong letters of reference, and get good scores on the GRE. :)
What sort of jobs (and at what universities and/or companies) do graduates in comp ling from UT typically get?
Both academic and industry. See our page about graduate placement.
How many papers are published every year by CL faculty and students at UT Austin?
We regularly publish at the main CL venues, including at ACL, EMNLP, IJCAI, EACL, and NAACL. Yuk Wah Wong and Ray Mooney even won best paper at ACL in 2007 and Pascal Denis and Jason Baldridge won best student paper at NAACL in 2007.
We seek to get new students up to speed as quickly as possible so that they can also submit their work to conferences by their second year.
I'm not sure if I want to get a masters or a PhD - do you have any advice about how to choose?
Our recommendation is to apply for the PhD if that is what you ultimately want. If
you get in and later decide to shift to a masters, its quite
straightforward. To go from masters to PhD will require a later faculty
vote after you've been in the program.
The linguistics department is currently very limited in the places it can offer to masters students, so we expect at most 1-2 students per year to be accepted for a masters with a computational linguistics focus.
I've completed my masters work elsewhere and have been developing my research in some detail already. Is it a possible to finish this research in the University of Texas under the supervision of any of the linguistics department faculty?
It isn't possible to go straight into dissertation writing straight away. You would need to apply through the usual channels, be admitted to the program, and take coursework before you could work on the thesis. If you feel we are a good match for you and don't mind doing coursework again, then you are certainly encouraged to apply.
In some cases, some courses can be skipped, but one can never skip the entire course sequence. One of the key aspects of taking courses is that professors get to know you and evaluate you before taking you on as a PhD student.
What is the timeline for studying CL in the Linguistics department?
You do 2-3 years of coursework. At the 3 year mark, you need to submit a qualifying paper, and if that is accepted you advance to PhD candidacy and can then be officially taken on as a student by a faculty member.
The qualifying paper is the major milestone. One also must take two seminar courses that involve writing a paper, and one of those must be outside your major area. There are no qualifying exams.
The linguistics department has recently changed its course requirements, leaving much more opportunity for students to take advanced CL seminars in our department as well as courses in other departments.
We expect students studying CL to finish in five to six years.
I think I'm interested because my undergrad linguistics minor was so interesting that I started considering a career in linguistics, but I also don't want to stop using my skills in computer science (my major).
You certainly don't have to give up any CS. Both Katrin and Jason have degrees from CS departments (Saarbrücken and Edinburgh), and their work is squarely within the realm of research topics that are pursued in CS departments and require the same amount of formal rigor and programming aptitude. Some of our courses are designed for linguistics who are learning their first steps in programming or on linguistically oriented topics in syntax and semantics, but the largely part of our graduate courses would fit well in a CS department. (At least one undergraduate course, Natural Language Processing, is already cross-listed as both Linguistics and CS.)