Masters in Science with NLP focus

[albert_hawking]

I am a (now) senior year undergrad at University of Mumbai and I want to pursue a grad program that is focussed on Natural Language Processing. I have two questions:

 

1. Which are the best schools currently for NLP research that have good masters programs and plenty of funding? There are other posts discussing this, but they're at least three years old. Does anyone have new information?

 

2. I don't know what to make of my profile, but I'm trying to dig for schools I could reach for. I consistently scored a 55% in all of my first four semesters here and then 67% in both my fifth and sixth (First Class, but about 3% less than a First class with distinction). (As a fresh undergrad, I was deluded and made to believe that course knowledge is useless, and my focus should lay purely on coding. I grew up and two years later, realized how severely limiting my lack of knowledge was.) Anyway, I'm in fire-fighting mode for the last six months, when I found my passion for NLP, taking up MOOC courses to bridge the gaps. (So far only Algorithms at Stanford)

 

My GRE score: 320 (160 Q, 160 V, 4.5 AWA)

 

Right now, I'm in the middle of my dream summer research internship (2 months) at the prestigious Indian Institute of Technology  (IIT-Bombay) under Dr. Pushpak Bhattacharyya who is the current vice president elect of the ACL (Biggest name in NLP in India). 

 

The MLT program at CMU is my favourite choice so far. Contacting them, it seems they find my GRE score good, and they've assured me that my profile will be judged in its entirety despite my academics. That still doesn't mean I have a good shot.

 

What is the best course of action for me? Must I wait out an year and still improve my academics this year? 

Furthermore, my aim is to pursue a career in research. I think the MLT program will be the best way to toy with the idea of a Ph.D But I know it's a long shot, Im willing to take detours to my eventual goal. Thoughts?

 

Sorry about the long post. But I really couldn't have done better.

[albert_hawking]

Is there anyone who can provide some insight and advice?

[DeleteMePlease]

Well I don't know how to convert your GPA... maybe do it for us?

 

Your GRE looks good. I think it is more important than your GPA for some schools if you're an international applicant. Also, the research internship will help a lot.

 

Definitely check for schools that don't have a minimum GPA cutoff. Problem is most of them have one but don't say so. What you can do is contact the profs you're interested in before you apply. If possible, ask them some in-depth question about their research, this means you should read their papers and show that you understand what they are doing and are interested in it. If you're lucky they will remember you when reviewing applications. 

 

Also, ask the supervisor of your internship or some of your undergrad professors if they have connections to professors in the states (i suppose that is where you want to go).

[albert_hawking]

Well I don't know how to convert your GPA... maybe do it for us?

 

 

 

Yes, this is about US schools.

I'm not completely sure, but my aggregate should be around 59% (that comes to a paltry 3.0 - 3.2 ish). I've scored badly in key subjects like Algorithms, Data structures as well. Which is why I'm trying my best to steer myself away from this potential career suicide. I'm not even looking to pursue a Ph.D. at this stage since my profile is hardly competitive.

A good 2 years at Grad school should help change my case for the better, but I'm worried about losing out on the good masters programs as well. Would it make better sense to wait out one more year work as an intern or research assistant and then apply with a better GPA (using my 7th and 8th sem scores as well) ? Or should I settle with a lack lustre masters program if I don't get the preferred ones? How must I decide? 

Thank you so much for taking the time to help me

Edited June 22, 2014 by pikachu!

[DeleteMePlease]

Well, A 3.0ish GPA is not good. But it's also not horrible. 

 

Hard cutoffs are usually at 3.0 for Masters as far as I know. Maybe don't aim for top schools. If you finish your Masters with a good GPA (and research experience!) at a lesser ranked school, you still have good chances to get into awesome programs for your Ph.D.

 

If money isn't tight you can just apply to a broad range of programs. I am pretty sure there are also less prestigious schools in your field that do awesome research.

[albert_hawking]

Indeed, I am considering the masters option and researching schools. Would you happen to know any schools that do good research in NLP that I might consider applying? 

[PhDerp]

I know many schools that are generally good for natural language processing (MIT, Stanford, UC Berkeley, CMU, USC, etc.)... but do you know what you want to do? What about NLP interests you? Is it data mining? Is it language technologies in general? Would you also consider yourself interested in speech technologies, not just text ones?

 

I bounced from NLP to speech processing, but applied to universities strong in both just in case.

Edited July 6, 2014 by PhDerp

[albert_hawking]

Hi PhDerp,

 I like language technologies in general. I'm much inclined towards sentiment analysis, but have had little exposure except for my junior thesis. On the other hand, my current internship has working in machine translation and I'm learning quite a lot. I like that too.

 

Could you help me with some resources which could help me gain basic literacy with regard to speech processing? I've not yet delved deeply in that area. Perhaps it could take me a few more weeks of research before I can narrow it down.

I have three more questions (sorry!)

 

1. Is my GRE score good enough to apply to a Ph.D or even a program like MLT at CMU?

2. Just how strongly is GPA weighed? Is there something that I can do to offset my poor performance in the first two years at college?
3. What universities are best for Ph.D/ MS programs in speech? How about NLP? (Please accept my apologies, that's four, but all information about this on gradcafe is at least two years old now)