Learn how people in your field
communicate with each other.
The primary mode of communication in Computer Science is the conference
paper. You may have already picked up some CS conference papers and
realized that they're a lot harder to read than your textbooks. Why is that?
The reason is because conference papers generally have very strict limits in
page length, and thus the authors are forced to write them in a very dense way.
Often times, journal papers will be much easier to read than conference papers,
because journal papers have longer page lengths and are polished and edited a
lot more for clarity. Unfortunately, journals are not nearly as popular in CS
as conferences are -- by the time something is published in a journal, it's no
longer considered cutting-edge research in CS. So as long CS is dominated
by conferences, you must get good at reading this brief publication format.
The second mode of communication in
Computer Science is the talk. This should come as no surprise, given the
importance of conferences in CS. When people go to conferences, they listen to
talks. So to be a successful researcher, you need to get good at talking, and
watching how other people talk about their research. If you go to a university
where you are lucky enough to have a lot of visitors come to give talks, make
sure you take advantage of that by going to as many talks as possible.
(1) Figure out what the top 3
conferences are in an area of CS that interests you are, and start reading
papers from those conferences that were published in the last few years. If you
don't know what those are, you can probably easily find out by asking around.
Look for common themes. Is a specific topic mentioned in a lot of papers
recently? What are the different sub-areas of the conference, and what
researchers publish in those sub-areas? It is good to get a sense of who is
working on what, and the different territories that researchers have carved out
for themselves, since this will also give you some idea where you might want to
go to graduate school (I assume that is on your agenda).
(2) If at all possible, try to actually
attend some of these conferences. Yes, this can be expensive, but there are
ways to make it more affordable if you plan in advance. The student rate for
the conference is usually substantially cheaper than for anyone else, and often
times students can even get free admission if they sign-up to do some
volunteering. There are also various travel grants you may be eligible for,
either provided by the conference itself or provided by the school you attend.
Keep in mind that anyone can attend conferences, and you don't have to actually
have a paper published in the conference to go. Attending a conference will
give you a lot of bang-for-your-buck, because not only will you learn about the
most recent research in the area you're interested in, but it's often easier to
digest the research trends from being at a conference in person than just
reading the papers alone. You will walk away with a lot of research ideas. You
will also get to associate actual faces with names you've heard of, which will
be good for networking and figuring out who you might want to work with in the
future.
(3) Keep in mind that usually the top
conferences in any given area of CS are very broad, and cover a lot of research
topics. Because these conferences are extremely competitive, often times only a
small number of papers from each topic will actually get published there. If
you find that a particular topic is over-represented and has a lot of papers
published in it in a given year about that topic, it means that this subarea is
very hot and fashionable to do research in at the moment. Conversely, don't
assume that just because a particular topic does not have many papers published
in a top conference, that this means there is very little research being done
on this topic. There may be a ton of research on that topic, but it may have
splintered off into more specialized conferences that have their own
communities. So if you find specific topics that interest you, that is narrower
in focus than the top conferences of the general research area, then you should
look at specialized conferences for those specific topics as well. This will
give you a better sense of what research is out there, what the bar is for
research on a particular topic, what has been done, and what has not been done.
Also, try to ask yourself why particular papers on a given topic might end up
in a top conference, and others don't? Do you see a difference?
No comments:
Post a Comment