BKMD
-
Posts
101 -
Joined
-
Last visited
Posts posted by BKMD
-
-
A recent PhD graduate of the Stanford NLP group suggested UMass to me, but I can't find much in the way of NLP/CL, at least not online. Anyone have an idea about that?
For NLP specifically, UMass has Aria Haghighi (student of Dan Klein) and David Smith (student of Jason Eisner), but they also have a ton of prolific machine learning people who often focus on text, including Andrew McCallum and more recently Hanna Wallach.
I don't think anyone has mentioned it, but Edinburgh is also famous for NLP.
As for qualtiy vs quantity... I do think there is a benefit in choosing a school that has more than one person you could see as an advisor, since you never know if a professor will be there for a full 5 years, plus it's hard to know how an advisor will work out until you actually start working with them. Of course quality is more important. I think every school that's been mentioned so far has some very high-quality people.
-
I believe the schools with the biggest NLP programs are (in order of size) CMU, JHU, and USC (ISI).
Berkeley and Columbia don't have many people but they have very good people (Dan Klein at UCB, Julia Hershberg and Michael Collins at Columbia). Stanford also has a good group (Jurafsky and Manning wrote what I believe are the two most commonly used NLP textbooks).
Most of the other schools in your list don't have many big NLP people that I know of. I know Irvine has a good data mining group, and Princeton has David Blei who is also very big machine learning and often works in text applications. I don't know as much about the other schools.
-
I'll add to this discussion that the size of the group might not be as big of a factor as the people who are in it. Some important questions to consider: how often can you meet with your advisor individually (and how useful are the discussions? this can certainly vary). Does the entire lab meet as a group on a regular basis? (I find these meetings very useful; not all labs do this.) Are there other opportunities to interact with the advisor? How well do the members of the group collaborate? Do the students teach each other and learn from each other? Do you personally hit it off with the advisor and the other students? These are the people you'll be learning from and working with, so how well you interact with them and enjoy their company is actually an important factor.
I also find that you can often learn a lot from other labs and departments, since after a while your immediate circle will have already taught you everything they know, so this is another thing to look for.
If you have an opportunity to visit the schools you're choosing from, you definitely should.
-
I'm at JHU and while I don't know anything about ISI, I can say that besides healthcare, the school also has close connections to government/defense careers (DoD, NSA, etc), probably from being located in the D.C. area. Many students have security clearances and are funded in this way.
-
I would withdraw from schools that you haven't yet heard from, because waitlists are actually not very common, and most likely you will deprive someone of a spot for no reason. You can even explain that this is why you are withdrawing early, and if they ask why you chose a different school (they probably will) they will understand that your wife's location was a factor so don't worry about offending them.
For the other schools, I would say go for it - it's definitely a good experience to visit and network. (Though I wouldn't be upfront that you aren't interested in actually attending because I know for a fact that faculty get annoyed when they fly out people who already know they won't accept an offer.)
Last year I withdrew from my safeties once I got into a top choice. I ultimately ended up choosing a different school than what I thought I would, so I'm glad I made the visits that I did. I think this is the right strategy: don't withdraw an app unless you're 100% certain, but if you ARE 100% certain, then you should withdraw/decide right away for the sake of others.
-
Another thing you should do when deciding is look at the alumni webpages from the schools and see what kinds of jobs their graduates have gotten. (And if they don't give this information then that's probably not a good sign.) You might also find that some schools tend to send more people into industry while others send people into academia, so you should figure out which kinds of jobs the schools are connected to and choose the one that fits with your career goals.
-
Without any other information, I think higher ranked schools are more likely to give you a better education and job prospects (that's why they have that ranking, after all), but I also think there are a lot of exceptions and it really depends on your specific research area. Last year I had to choose between a #1 school and a #30 school and I went with the lower one because I honestly felt like I would have better opportunities there, and I still feel like that's true. There are a lot of factors to consider.
The advise I got from professors when I was making my decision was that the most important factor is who your advisor will be. If your advisor is well-connected, then having their recommendation can do more to get you a job than the name of your school. Keep in mind that there are dozens of extremely good professors in any research community, yet generally any one school will only have one or two of those professors, so there are certainly top professors at lower-ranked schools (but which schools depends on your field - I'm not saying every lower-ranked school is just as good).
I'm told that the most important qualification when applying for PhD-level jobs is your publication record, so it's important to go to the school where you think you'll be most productive. This means having a good advisor as well as being in a collaborative environment, the ability to push your own ideas forward, etc. Of course, one difference that the ranking can make in applying for jobs is that applicants from lesser-known schools might be filtered out before the application is even considered. As long as your school is well-regarded enough that people will look at your application, having an exceptionally strong CV should have a stronger impact than having an okay CV from a top school.
Someone mentioned that teaching experience is important for academic jobs, yet can be hard to get. That's another reason I chose the school I did - they let you teach your own courses, which is much better experience than being a TA. Try to find out what perks and opportunities are available at whatever schools you end up choosing between.
- nhyn, Strangefox and Eigen
-
3
-
Last year I got a personal email informing me of my CMU acceptance, but the official notification didn't go out until a week later. So don't get discouraged just because some people have gotten some initial feedback. Also, the different departments all give decisions at pretty different times.
-
The general wisdom is that they don't care unless you do exceptionally poorly. I'm not sure why some schools even still require the GRE.
-
I remember last year UIUC sent out admission decisions in several batches, so don't worry if you haven't heard anything.
-
Congrats on the interview. I'll just agree with what everyone else has said in that you should definitely go, because ultimately what matters in your choice is the people that you'll be working with (your advisor(s) and lab mates) and you won't know what to expect until you meet them. I had similar conceptions about the elite school I ended up attending but I was completely wrong. I think this is partly because engineering is generally unpretentious, but there's also the fact that grad students are being funded rather than paying their own way, so there are fewer class barriers.
-
I went to a recruiting weekend last year and I should warn you that it was very intense and exhausting. You might want to do as much studying for your midterm as you can in advance. However, it was well worth it because it ended up changing my perception of the school, and I ended up going there even though I hadn't expected it to be my top choice going in. It's important to visit every school/lab you get into because the people you'll be interacting with are really the most important factor, so you should try to meet them. (If they're interviewing you, it means you're on their shortlist, and it sounds like they want you.)
From your last message, it sounds like you've decided to go, so that's good. As for dress - I would ask one of the students or administrators, especially if there is a person coordinating the events. No one will mind you asking, and it's less awkward than asking faculty.
-
I don't know enough about the admissions process to say what your chances are, but it seems like you have a decent chance. I know they usually look for research experience so this is a drawback, but your exceptional GPA and the fact that you have a masters should help a lot. They probably do need to see evidence of research ability so I think you'll need to make a strong effort to convince them of this in your essays - try to come up with somewhat specific research ideas and how you would solve them, and be specific about the professors you're interested in at each school.
Computer vision in California... I think UC Irvine is good. Stanford/Berkeley as well but of course they are hard to get into.
-
It is actually pretty common to go from a Masters to a PhD, and at a lot of top schools, they sometimes admit PhD applicants into the MS program (not necessarily funded), with the option to transfer if they succeed. Some schools favor PhD applicants who already have a Masters and some don't, so it's hard to generalize about this, but I wouldn't worry about this.
I would be clear in your essay that you're interested in doing research and might be interested in eventually pursuing a PhD. Enough schools offer research-oriented MS degrees that you should be able to find an option that works for you. What I wouldn't do is apply for a PhD program with the intention of dropping out after 2 years, because that's essentially cheating the school into giving you a free Masters and won't look good. Just be honest about your goals and I think your situation is fine (and not actually too uncommon).
-
Oh yeah I was going to make suggestions for schools... Kansas State is good for text/data mining and you'll probably get in. Some more competitive schools but maybe in your range that I think are good in this area are U Pittsburgh, UC Irvine, and UT Dallas.
-
I think your research area is actually more popular than you think. My understanding is that text mining is really big these days, with tons of money coming from both government and industry. Don't worry about this.
To me your application looks pretty decent but not great. You have some research experience and you seem sincerely interested in it, so it will help if you let that come across in your SOP. My main worry is your GPA - while it will help that you've improved in the last 2 years (and most schools will say they give much more weight to these last 2), it's possible that your app will get filtered out before they even look at it just because the overall GPA is on the low side (though at least it's not below 3, so you'll be okay at a lot of schools).
I don't think you'll get into a top school, but you never know. Pick your favorite and apply. JHU is nice because it has a low application fee ($25).
I would aim for mid-range schools. BTW is there a reason you're only considering schools in the northeast?
The best thing you can probably do at this point is to just have a strong and focused essay. It helps if you have specific research ideas with citations because it shows you know what you're talking about. Don't do what you suggested and be shy about your interests, be enthusiastic and name specific faculty that you think you'd be a good match with, to show that you're interested in their department. Generally to get admitted, there needs to be at least one professor who wants to advise you, so be sure to find the right schools for your area (but like I said, it's actually popular, at least at the bigger schools, so this should be okay). Good luck!
-
I think they are probably comparable in reputation. If you plan on getting into serious research or going on to a PhD, then the main factor should be the faculty you'll work with and their research. If not, then probably the main factor is the job opportunities you'll get and how much you'll get for your money (compare the tradeoff of tuition+living costs vs what kind of job you think you can get). One thing that might help with this is that sometimes departments list their alumni along with the careers they had upon graduation, which can help you gauge what to expect.
-
Yeah, it sounds like you have a very good application, but you probably need to demonstrate that you know the core CS material before you can start such a program. There is a lot that is important to know - if you still have a chance to take the subject GRE, it would be a good idea...
-
Any clue on the California thing?
My understanding (someone correct me if I'm wrong), there is a tuition difference between residents of California and non-residents (I think this is true of any state school), but I don't see that there would be a tuition difference between American students from other students and international students. And since most of the American students come from other states anyway, there is probably not a difference in the way they perceive domestic vs foreign. I could be wrong about this but that's my impression. Berkeley's budget situation is perhaps worse than many other public schools, so this might be an bigger factor here. I would apply if you want to (though keep in mind, the application fee is a bit expensive, and admission here is definitely very very competitive).
Overall, I would say your test scores are very good. I don't think they will necessarily help you, but they are good, so I wouldn't worry about that. I think you have a decent chance at top schools but you should apply around, because your application doesn't seem exceptional (it's good and you should apply to good schools, but I don't think it stands out enough for #1 schools).
-
Yeah, I think having a first-author publication in a recognized conference will help more than anything.
The main reason they might care about which school you went to is so that they have an idea of how well you might know the CS material. If you are concerned that they won't be familiar with your school, you can offset this by doing well on the CS subject GRE. If you do well, then this will give the admissions committees confidence that you are well prepared. This exam is pretty hard and requires some preparation, and it's pretty soon (I think they offer it in October and November), and I'm not sure how easy it would be for you to take where you live. But if it's an option and if you think you would do well, it would help. (though I wouldn't send your scores to schools unless you do well, say >80%).
Normally I tell people that the subject GRE is a waste of money, but I do think it can help if you're coming from a school that isn't well-known in the US.
-
Hi,
It sounds like you have a great profile, so I would encourage you to apply to top 10 schools. I agree that UIUC is good for systems (though I don't know much about this area so I can't recommend anything else). Of course it's always hard to predict, but it's worth applying wherever you are interested - you have a decent shot.
-
Personally I didn't contact anyone but it sounds like it's a good idea. The most important thing is to list who you want to work with when you write your admission essay - near the top so that it's noticeable - and usually your application will be forwarded to the right people.
-
Some schools that I think are great in this area but have less competitive admission: UC Irvine and U. Rochester.
-
Like the other person said, they don't often care about the GRE, though as an international student, they might check your Verbal score - you have a good score, so you're in a good position there.
It seems like you have a decent chance for admission, but it's always hard to say. You still have a few months before you apply, so I think the best thing you can do in that time is 1) stay close to whoever will be writing your recommendation letters, because it is pretty important that they can say specific things about you, and 2) do more research if you have the opportunity (even if you have do it on your own).
NLP schools
in Computer Science
Posted
I can agree with that. I feel like I get more from working with a variety of people, since there's only so much one person can teach you.