2

It isn't definitely a "linguistic" question, but I hope there're some people who solved the same problems like me.

Since of my childhood I was very curious in languages and linguistics, but I hadn't enough money to study abroad and universities with good linguistics faculties. Finally I was graduating in computer programming and worked in software engeneering since I was graduated.

I want to ffind a way to connect both of my interests, but I don't know anyone who tried it, and I'm not sure about current situation in this field. For example, as long as I know, the spellchecking and parsing the search requests are well-developed enough and these problems are almost solved.

So, only a great research or a new university degree can help me with finding my place in that field. But is it possible from online courses, or I have to join "full-time" university (the second is impossible, I think :(). The thing is I don't know anyone who worked with such a problems (all of the linguists I know studied culture, histore, etc of the countries).

  • 1
    I'm not sure I understand what your exact question is, but I'll try anyway: You already specified computational linguistics as a field where computer science and linguistics meet, another keyword is Natural Language Processing. As for online courses, these are available, provided you don't want/need the formal qualification of a degree. You'll need to specify what kind of online course you want exactly. – robert Oct 9 '13 at 10:06
  • 1
    And btw, parsing is not "solved". Parts of speech tagging is relatively good for common languages (but not perfect), and unsupervised/unchecked parsing is quite inaccurate (which is no say there hasn't been tremendous progress). Many other problems, such as automatic semantic analysis, remain. – robert Oct 9 '13 at 10:11
  • I'm personally not satisfied that POS tagging is solved. It can hide ambiguities from subsequent stages and it can be hampered in the same way by being isolated from an imperfect word breaking stage before it, especially for languages such as Burmese, Chinese, Japanese, Lao, Khmer, Thai, Tibetan, and Vietnamese. – hippietrail Oct 9 '13 at 11:03
  • What specficially interests you in linguistics? Do you want to work in Natural Language Processing? Are you interested in foreign languages and how the relate to each other? – Christian Oct 9 '13 at 11:44
  • 1
    You may find these two chapters on different ways to implement your ambitions useful. Both were written (in 1998) by linguists who were also programmers. The techniques have gotten much better since then, but the approaches are much the same, and the points the papers make about the differences are still valid. – jlawler Oct 9 '13 at 15:45
4

If you are still a student (or if there is a university in your city), I suggest emailing or making an appointment to talk to some professors specializing in AI. Preferably ones specializing in Natural Language Processing (NLP) or Computational Linguistics (CL) but even if they don't they can still give you relevant advice and point you to the proper people.

Other than that, you can take an online course. Coursera.org is a website devoted to providing free access to courses written by professors from world-leading institutions. Currently they have not one, but two introductory NLP/CL courses. One from Standford University and one from Columbia University. The Stanford class in particular was co-created by Dan Jurafsky who also co-wrote the textbook that most introductory NLP/CL courses use. Just click the "preview" button on those courses' websites and you should have access to all the slides and video lectures in the course.

At the time of writing it appears that the Columbia course is about to begin a 10 week long session, meaning the materials are unavailable for preview. However, if you register for the course you will get access to the materials as the course progresses. Not only that but you will get to participate in weekly assignments, get feedback from other users and TAs, and at the end get a nice certificate of completion which, if nothing else, shows employers and potential professors that you're serious about your studies.

These courses should give you an overview of NLP/CL tasks and techniques. You can use this to guide any further research.

I would also caution you that linguistics is not learning and speaking languages. It is the study of the underlying mechanics of language which all languages share. I'd start by reading The Language Instinct by Steven Pinker. It's very basic and written for a general audience but it provides a nice overview. If you enjoy that and want to learn more then MIT OpenCourseWare has slides for MIT's introductory phonology, syntax, and semantics courses. With that foundation you can explore all the Linguistics courses MIT offers and see if any interest you.

| improve this answer | |
1

NLTK is a NLP library for Python, and the NLTK Book covers many computational linguistics topics.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.