How to Apply

Preparing for Graduate Study

(work in progress)

Here are my advices on how you should prepare for graduate study in NLP. While these are what "I" consider to be important, I believe they are generally important for most NLP and machine learning labs in or outside of Korea as well.

English

Most written communication in our lab will be in English (spoken communication can be in Korean if you prefer). During graduate study, you will need to read and write a lot in English. Most students especially struggle with writing (than reading). It doesn't mean you need to know a lot of rare words or use a fancy syntax. You just need to be logical, succinct, and clear. Often times writing can be improved by reading. Note that academic writing is quite different from newspapers. I generally recommend aiming for a high score in TOEIC or TOEFL if you need a concrete goal to work on. Reading NLP papers will also help. See NLP below for what papers to read.

Coding

If English is the universal medium for communicating with other "people", code is the universal medium for communicating with "computers". Unless you are strongly theory-oriented (where math might be the most important medium), you will need to code a lot during your graduate study. Modern AI programming is mostly based on Python 3, so you need to be familiar with it. Think of it this way: if you dedicate a few weeks of Python 3 programming now, you will save time equivalent to a few months during your study (and years in your life). So it is a very good investment.

Suggested materials

LeetCode: you should be able to solve most easy problems. Work on them and practice. A similar-level question might be asked during the interview. But don't put too much time on it. In particular, I personally don't recommend working on hard problems. Rather, work on other things in this page.

Coursera Python 3 Programming: Be sure to know most materials covered in this class. For instance, it is useful to know that you can append two lists with '+' operator. It will speed up your coding a lot.

Computer Science

Some basic knowledge in computer science will help you throughout your career.

Math & Machine Learning

Mostly about machine learning fundamentals. This also means you have good foundation in mathematics.

NLP

This is mostly about knowing the state of the art.

BERT

In modern NLP, BERT is the fundamental building block.

Project Experience

Now you know what you need to know, it is time to make a contribution!

Research Experience

For MS (석사) admission, I do not strictly require previous research experience. For MS/PhD Integrated (석박통합) and PhD (박사) admission, I do expect some research experience.