2/28

Hello hello everyone and welcome to my fourth networking interview! This time, I was able to interview 신효필 교수님 or Professor Hyo-Pil Shin, a professor of computational linguistics and NLP at Seoul National University!

Similarly to my previous interviewee Wonik Cho, I found out about Professor Shin through research he’d done previously. In 2018, he and another researcher published a paper titled ‘Grapheme-level Awareness in Word Embeddings for Morphological Rich Languages'” where they talked about how to analyze morphologically rich languages, like Korean, for the purpose of word embeddings. For those of you that don’t study linguistics, I’m sure this makes little to no sense but this was extremely interesting research for me to read, especially as it pertained to my research at the time. So with that commonality in mind, I reached out to Professor Shin for an interview and he quickly agreed!

Since it was our first time meeting, we briefly introduced ourselves at the beginning and then sat down for the interview. I asked him to tell me about his educational background first and I was pleasantly surprised to find out that we had something else in common (a distant commonality but one nonetheless). Professor Shin attended Seoul National University for his undergraduate and graduate degree, as well as his PhD, where he studied Linguistics for all three degrees. In 1995, he came to the States and began his second Masters degree at the University of Missouri, or Mizzou. I also have ties to Mizzou, in that I applied there during college applications back in high school…that’s about it haha. Great school though! During his time at Mizzou, Professor Shin studied Computer Science, which was not his original path, as we saw previously. When I asked him about why he studied Computer Science instead of Linguistics, he told me a bit about a barrier he faced in relation to studying in the States.

In the 90s, getting admission into a university in the States as an international student was much harder than it is nowadays. Many international students study full time in the states today but when he was applying, it was much more difficult, especially if you weren’t studying in STEM related fields. There was a desire for computer science, engineering and other stem backgrounds for schools in the States at the time. So, with that in mind, Professor Shin decided to start studying computer science, which was a very smart decision on his part as the computational linguistics field was really beginning to take off. After receiving his final degree from Mizzou, he immediately began working at New Mexico State University as a senior researcher in their Computing Research Lab. At NMSU, he was a specialist NLP researcher and he fondly reminisced about those three years, saying that he thoroughly enjoyed the three years he spent there.

After his time at NMSU, Professor Shin spent a year at YY Technologies in Silicon Valley as a senior researcher. When he started in 2001, YY Technologies was a start up in California that was working on programming automatic email response systems, which weren’t common in the early 2000s. So he and the company were working on a technological contribution that would be extremely important to us today, one that we often use without considering the work that went into it so many years ago.

But eventually, Professor Shin wanted to come back to Korea, and specifically, back to Seoul National University. So in late 2001, he returned to SNU as an Assistant Professor in the Electrical Engineering department where he taught programming and NLP classes. After 2 years he returned to his roots, becoming a Professor in the linguistics department where he teaches computational linguistics courses and other intro linguistics courses today. When I asked if I would be able to take one of his lectures, Professor Shin said that his courses are particularly popular so it may be hard. That and they require computer science skills, of which I have none haha, so I sadly won’t be able to take any of his lectures.Professor Shin was also the Associate Dean of Academic Affairs for the College of Humanities for 2 years and worked at SNU’s Graduate School of Data Science for 2 years as a professor and the Associate Dean of Academic Affairs.

After hearing about his academic and professional history, I asked him about what interests led him to this position (quite out of order, I know). Professor Shin said that from a young age, he always wanted to be a scholar or a professor, but he didn’t know exactly what he wanted to study or teach. But he did have a love for and interest in languages, so he took the obvious route and began studying linguistics. He started with everyone’s favorite: syntactic phenomenons and semantics. And like so many of us, once he started studying he was hooked and hasn’t left the field since.  Since I am also a linguistics major, we briefly discussed our thoughts on linguistics and we found at that we both disliked phonetics and phonology, lamenting over how difficult the subject is. We also learned that he worked a bit on translation technology, specifically rule constraint based models, which are a much older model type compared to the current NMT and Transformer based translation models.

After this, we spent some time talking about our respective research, with me asking many questions about the research of his I’d read. Since I don’t have a strong background in computer science, he had to explain some things to me and he suggested that learning computer science would be helpful to me. As much as I am deathly afraid of computer science, I have to agree with him. Since I’m interested in the ever growing computational linguistics field, it only makes sense to learn a huge component of the field if I ever want to contribute to it. So you may see me in one of Wellesley’s impossible CompSci lectures one day…maybe ㅎㅎㅎ.

I also asked Professor Shin about any current research he’s doing. He told me about how he and a team or graduate level research students are currently working on building a transformer based pre-trained model for the translation of law and medical documents. I found this particularly interesting as I had yet to read many papers on Transformer based translation models, despite their growing prevalence in the field. So I’m very excited to see how the creation of this model works in translating such important documentation like legal and medical documents.

Finally, I asked Professor Shin the long awaited question: Do you have any advice for students who want to become professors or for those who want to do research/remain in academia. For the question about becoming a professor, his first answer was the typical “Study Hard” answer. But his second answer was quite interesting.

If you want to become a professor easily, you need to study AI, Computational Linguistics, Corpus Linguistics, etc.” 

When I asked him to explain, he said that becoming a professor is much harder nowadays, especially for humanities fields, so teaching in a field that is partially stem would be very beneficial to me. He said that purely theoretical fields like Humanities and Social Sciences can be hard to work in. He followed it up by saying that it was a half joke, but I think that Professor Shin is correct. Becoming a professor is very difficult and if you’re teaching in a less sought after field, it’s even harder, so improving my chances by expanding my skill set would only benefit me.

As for my question about someone who wants to remain in research or academia, he was a bit more positive, saying:

Study what you’re interested in, what you can do your best in. Concentrate on what you want to do. I suggested that you study Computer Science but if you aren’t interested in it, don’t do it. Do what you want to do

I really appreciated this advice. As I’ve said in previous write-ups, I’ve been worrying about what I want to do and how feasible/long term my ideas are, so hearing someone essentially say “Do whatever you want” felt really nice, even though he said before that I have to choose my field wisely haha.

This interview write-up seems a bit shorter than my other interviews now that it’s all written out. I think this is because we spent quite a bit of time talking about our respective research and the current state of translation technology and computational linguistics. And as much as I’d love to write all about that topic, I think it would just confuse everyone haha. But despite the shorter length, I was still able to learn quite a lot from Professor Shin. My other interviews had been heavily focused on different career paths with only a bit of talk about research. But as I continue my undergraduate career, I feel myself being more drawn towards research, so having an interview where we talked in depth about this path really helped ground my interests a bit more.

For that, and everything else, I want to say thank you to Professor Shin. I really appreciated the opportunity to interview you and speak with you about your experiences and the things you’ve learned along the way. And I’m especially glad that I could interview since you haven’t been interviewed in a while haha. I really enjoyed meeting you and I’m wishing your lab the best of luck with your model!

Thank you very much for reading! My next (and final) interview will be with 김성곤 교수님 (Professor Kim), a previous Dean at SNU and previous two-time president of LTI, a Korean translation institute, among many other careers. He is currently teaching at Dartmouth College so our interview was done over email, but this mode of communication did not prevent him from sending some of the most interesting and thought-provoking responses I’ve ever read. I’m very excited to share them with you all, so I hope you all will read them in my final interview write up!

Until then, 건강을 조심하고 맛있는걸 잘 드세용~~!