Parth Verma - University College London

Parth Verma (CS1) Domain: Quantitative Finance (Sentimental Analysis)

Interned at University College London Mode: Online Interviewed by: Ahilaan Saxena

Introduction

I am Parth Verma, I am from the Computer Science Department. I interned remotely under Professor Philp Treleaven of UCL (University College London) during summer vacations at the end of my 2nd Year. The internship was based on quantitative finance and mainly involved sentimental analysis of stocks, i.e., analysing the market sentiment about a stock with the help of related tweets. 

Motivation

I chose to intern in this field because of my background in Mathematics and Computer Science and the vast bundle of opportunities this field offers. I did not have much significant knowledge about the subject. Nor do the professors expect a vast skill set for 2nd-year interns. A good CG and academic achievements like a good JEE Rank and Olympiad qualifications are enough to secure a good research internship.

Application Process

Beginning in September, I started sending out a couple of cold emails daily to professors from CS institutes all around the world with my CV and explaining why they should pick me for the internship. An excellent place to find a list of the top CS institutions is csranking.com. Out of every hundred emails I sent, I received around 7-8 replies. Some of them even took more than a month. Every professor has a different process. Some provide a direct response, while others conduct an interview. Prof. Treleaven, however, did not take any interviews.

Internship Experience 

In December, Mr Treleaven finalised our team of four—three from IIT Delhi and one from IIT Roorkee. He provided us with reference material on sentiment analysis and asked us to source datasets related to stocks from the web. We acquired datasets for Amazon and Tesla stocks. We examined current methods and various libraries used for sentiment analysis. Our approach involved analysing tweets to determine whether consumers felt bullish or bearish about a stock by assigning scores to each word and performing calculations based on those scores using custom dictionaries. At one point, we noticed that the tweets indicated unusually bearish sentiment. To address this issue, we manually adjusted the scaling of the negative tweets. Although the internship was remote and did not provide a stipend, it was an incredibly enriching experience that offered valuable insights into Machine Learning and NLP (Natural Language Processing).

Getting an LOR or not largely depends on your relationship with the professor. You should request it; in most cases, they will be happy to provide it. However, if you haven't worked diligently during your internship, the professor may deny you a certificate of completion.

Advice to Juniors

Besides the process I described earlier, you could also apply for a project under a professor at IIT Delhi during the semester or winter break, which is relatively easy to get. Completing such a project makes securing a research internship abroad much easier due to the relevant experience you have gained from these projects. Additionally, including assignments and projects you have done as a part of IITD courses in your CV can also be beneficial.