Exchange Diaries  

By International Relations Cell

Outbound Diaries 2020

Mansi Sinha | Mathematics

National Research University: Higher School of Economics, Russia

Research Area: Natural Language Processing

Q1: Which university did you get, and what was your research topic?
A: The University that I got was National Research University: Higher School of Economics, Moscow. My research purpose was Natural Language Processing. This project was about machine translation using Natural Language Processing.

Q2: How did you apply for the FT?
A: I applied through direct mailing. I was not a pre-final year back then, so very few portals were open for me, as most of them are open for pre-final years only. I tried direct mailing. I did get many positive responses before this, but the main problem I faced was with the funding. They were like-” I will guide you, but I do not have funds for this.” I didn’t have enough resources, so I was looking for a professor who has the same research interest and is ready to fund it. It took me around 300-350 emails to get a perfect match.

Q3: Why do you think FT is essential?
A: I think there are a lot of benefits to ft. I would not call it a necessary thing to do, but it will be a plus point if you're interested in research. In my case, the project that I worked on, there are parts about it that are still ongoing. And we’re targeting a research paper eventually, once all of that is done. If you’re planning to do a Ph.D. or MS, a research paper will help a lot in your candidature. And the second thing is exposure. You get to interact with people of different cultures, which, in my case, did not happen sadly as the internship had to be done virtually (due to the COVID-19 pandemic). But this is something that I was looking forward to, and I hope to get this experience sometime later.

Q4: From where did you get the motivation, and how did it all start?
A: I started my first research project in my third semester, and that was because everybody was doing something. I wanted to try my hands on machine learning before anything else, so I took a project under Prof. Pawan Goel from the Computer Science and Engineering Department. That was my first project in natural language processing. With that, I started building my profile for research slowly. I realized that I would pursue this field further, after about 6-7 months of doing the project. I could not apply for FT at the end of my second year. So I did a research internship at the end of the second year at IIIT Bangalore. That helped me a lot to eventually get an FT. If you do your project along with your semester, you do not get involved in it a lot. If you spend your December or summer on campus, that definitely adds up. I spent my summer in a research internship at IIIT Bangalore. That project and the experience helped me build a research profile and based on which I did get a lot of good responses for my FT emailing. I did have an interview for this properly and all the profs were interested in my previous research work. If we do not have any prior project/research work, it is challenging to get a foreign internship, as the professors are keen to know if we have done our homework before applying.

Q5: How did you handle your preparation along with the academics?
A: I decided around August that I wanted to do FT in the next summer. For that, obviously, Whenever I was free, I looked up the QS rankings, picked one university, found ten professors from that specific college who had the same research interests as me, and then added them to my list (Google Sheet). This part of the process continued till about September. From October onwards, I started mailing the professors. One of the fundae that I received from my seniors was - do not send an email to a lot of professors from the same college simultaneously for two reasons: One, your mails might go into spam. Secondly, they do know each other. If some professor decides to shortlist you and then happens to come across your name from another professor, the chance is lost. Follow-up emails are also essential. Wait for a week and mail again if there is no response from their side.

Q6: What is the role of CGPA to secure an FT?
CG does play an important role in pursuing Masters. My CG wasn’t very high when I had applied (around 8.3). A CGPA above 8 is always safe. For PhDs, CG isn’t essential. But for Masters, it is (9+).

Q7: Can you tell us more about your project and the name of your professor?
I worked under Prof. Attila Kertesz, and the project was Mathematical operations in recurrent neural networks. I had to use NLP and create my GRU to calculate the solutions for primary-level word problems.

Q8: Did you do a project in the same domain in Kgp?
Not exactly. The projects I did in Kgp were based on convolutional neural networks.

Q9: How did your university manage everything in this confusion during the pandemic?
They provided me with the credentials to their university server. All the work I did is there on my professor’s server.

Q10: Did you face any difficulties during this entire process?
It was a huge mental blow, to be honest. My fingers were crossed regarding whether I would be able to continue my internship virtually or not, as many students had their internships rescinded by the universities. My professor was very supportive, and he went ahead with the project. There was an apparent mismatch of time-zones, and I could not clear my doubts whenever I wanted to. But that’s a problem every student had to overcome.

Q11: What was the mode of communication with the university during the project?
We contacted over emails. We had informal calls over Skype.

Q12: What advice would you like to give to the fellow students who aspire for an FT?
If you are going for direct emailing, do not lose patience. It takes time to get the perfect internship you want. It might be frustrating at times. But just hold on until you get what you want.