Part 3 - Road Trips and Decision-Making

By Mandy Korpusik '13

Most of my fellowship
applications were due in November, and most of  the grad school applications were due in
December, so I wasn't expecting to hear back for a few months. However, I
received my first acceptance from UIUC while I was still on winter break! Once
I returned to Olin for second semester senior year, I was also accepted into
Penn (with a fellowship!) and UT Austin. Professor Victor Zue from MIT's SLS
(Spoken Language Systems) group called me to schedule an in-person interview
the day before the EECS department would decide which students to admit. I
asked a couple friends to drive me to and from MIT, and then I had half-hour
conversations each with Professor Zue and research scientists Stephanie Seneff
and Jim Glass. They were all very friendly and talked about interesting
research projects their students are working on. Then I had a Skype call with
Professor Luke Zettlemoyer from UW, which I thought would be an interview, but
turned out to be UW accepting me! Finally, while I was in the Logan airport
waiting to board a flight to Baltimore for an interview weekend at Johns
Hopkins, I got my acceptance email from MIT. I was so excited I emailed all my
friends and - of course - called my mom.

Data Gathering

Johns Hopkins

My advisor recommended that I apply to Johns Hopkins
because it has a center for language and speech processing (CLSP). I visited there
in February, and the trip was quite stressful because I had ten interviews over
the course of two days. They provided plenty of gluten-free food for me and treated
us to delicious dinners both evenings (one night we ate hors d'oeuvres with
professors and students in a streetcar museum, and the next evening we ate dinner
with grad students at an Afghani restaurant), but there was only one
information session and no tours. Johns Hopkins probably has more professors
working on NLP research than any of the other schools I visited, and I knew a
lot of students who were really excited about the prospect of going to Johns



Two weekends later, I flew to Illinois for my UIUC
visit. The dates of this visit conflicted with UT Austin's, so I did not have
the opportunity to visit Austin. Since UIUC is a large public university, they
have a huge department and many resources. I was impressed, and I really liked
two NLP professors I met there. It was great to see how much the UIUC grad
students loved their work and were able to strike a good work-life balance. The
students in the lab had a friendly, joking relationship with each other, and it
seemed that everyone in the Midwest was extremely pleasant. I went on a driving
tour of Urbana-Champaign, and the second evening we ate dinner at a fun place
with an arcade. I played several arcade games for the first time with two women
I met at JHU, and we won matching snap bracelets.


The following weekend I visited MIT. It was an easy
trip because it was only a short drive from Olin. The first evening was a
welcome dinner hosted by the grad student association. The next day, over 40
professors gave 3-minute explanations of their research projects, and if they exceeded
the time limit, the moderator "gonged" them. The professors at MIT excelled at
showing fun, exciting demos and describing their work at a level that's easy to
understand. Whereas at UIUC I met with only three professors, at MIT I met with
seven. Most professors said MIT is a great place to come if you're not 100%
sure what you want to do, since there are so many amazing departments (i.e. CS,
linguistics, cognitive science, etc.), and it's not uncommon to switch
advisors/groups. I met again with Professors Victor Zue and Jim Glass from the
SLS group, and I had lunch and dinner with the group. After dinner, I visited
the dorm room of one of the grad students, and got ice cream with another grad


Visiting MIT

The last visit day was so much fun! I went on a tour
with my roommate Ramya and several friends of hers from Georgia Tech. We got to
see the campus, and then we walked to the MIT museum for a dinner party and
took the T to Harvard Square for deliciously rich dark hot chocolate. Ramya and
I finished the night by heading back to the hotel with another friend and
watching TV. I felt so happy after my MIT visit that I already knew I liked it.
Coincidentally, I had previously met the other two students admitted to the SLS
group on our JHU visit.


Visiting MIT

first weekend of spring break, I visited Penn and saw Philadelphia for the
first time. I liked Penn's campus, and I met with two excellent NLP professors.
Both the happy hour and dinner were extravagant, featuring delicious salmon at
happy hour, and a four-course meal for dinner! The grad students were friendly,
and I made a friend from LA who swims 10 hours a week, goes to bed at 9:30 am,
and is interested in computer architecture. I spent time with my two friends
from JHU and UIUC too. The Philadelphia tour the next day was fun, but cold. We
experienced snow, hail, and thunder and lightning all in that one visit. I bought
dark fudge at the Reading Terminal and did my "Rocky" impersonation, running up
the steps at the Museum of Art. Philadelphia also had many wild partiers
celebrating St. Patrick's Day (of course).


Museum of Art - Philadelphia

After spending St. Patrick's Day back
at Olin and eating hotpot with friends during spring break, I traveled for 13
hours to Seattle for my UW visit (I accidentally booked my flight so late that
it was expensive and had long layovers in both directions). By this time, I was
starting to feel exhausted from all my visits, but I still took the opportunity
to watch online NLP lectures and read an AI textbook on my flights and while sitting
at the gates! Seattle and UW's campus were gorgeous. I got to eat a Greek lunch
on The Ave, visit a "millionaire" professor's house with an amazing view,
experience Seattle's unpredictable rain (I got drenched - again), see Pike
Place Market, and meet up with my best friend from high school for Indian food,
a tour of Seattle U, and ice cream. UW also has many professors doing cool
research related to NLP, so at this point I was definitely realizing this would
be a tough decision. 


Pike's Place - It's "touristy" to take pictures with the pig!

Making a Decision

It is important to know
what matters most to you when deciding on a grad school, and for me, I knew I
was looking for a school where I felt comfortable, the grad students seemed happy,
and the gender balance was similar to Olin's.

Visiting the grad
schools was a huge help. The schools worked hard to impress us, and reimbursed most
of our travel expenses. They booked us in expensive hotels and gave us goody
bags with free T-shirts and folders of information, and treated us to elegant
dinners. I had the opportunity to meet with many professors who are famous in
their fields, and I was exposed to cutting-edge research projects. Since many
of us visited several of the same schools together, I made a few good friends
whom I still keep in touch with. Seeing the schools and meeting the professors,
grad students, and admitted students gave me a good feel for each place, which definitely
helped me decide.

The form of financial
support offered by a grad school is also a factor in the decision. For example,
since Penn offered me a fellowship, I would not have had to worry where my
funding would come from and could choose any advisor without worrying whether
they had enough funding. The same is true for receiving an external fellowship,
but some fellowships may require recipients to work at specific companies for a
couple summers. One school could only guarantee funding for the first three
years, and the linguistics professor I talked to said she would have to
co-advise me with someone who had more funding. Sometimes, schools offer a
research assistantship (RA), where the funding comes from the professor you do
research with. Some schools may offer a teaching assistantship (TA), in which
you are expected to spend 20 hours a week teaching.

Since choosing an
advisor is so important and is often the reason grad students told me they
decided on a school, there are several factors that should be considered when
thinking about a potential advisor. First, the school you choose should have
several potential advisors, in case you have an issue with your first advisor. I
talked to several grad students who switched advisors, and most schools are
supportive of this. A potential advisor should be doing research that interests
you and should be a good personality fit (different people have different
opinions on which of these two is most important). You should also consider the
potential advisor's age and whether or not he/she has tenure. Older advisors
have more experience and probably have tenure already, whereas young advisors
are often more motivated to get their students to publish because they are
working toward tenure. I would feel somewhat nervous working with a professor
who is not tenured because I had seen two examples of professors switching to
different universities and how it affected the students they left behind.

 In the end, I
decided on MIT because I felt the most comfortable there and knew exactly which
group and professor I would work with. Although other schools described MIT as
a "pressure cooker," I did not get that impression at all. In fact, the grad
students in the SLS group seemed really happy, everyone I met was energetic and
enthusiastic, and the students seem to have fun and participate in many extracurriculars.
In addition, MIT is surrounded by companies doing language processing (i.e.
Nuance, Apple, Google, Amazon, and Microsoft), they guarantee financial support
(usually Research Assistantships),
and the requirements aren't bad. Plus I could live in a graduate dorm, and
there's a swimming pool in the CSAIL building. UW was a close second choice,
but I just couldn't turn down MIT.


 Final Decision- MIT

Posted in: Graduate School, Scholarships and Fellowships