We often use Naver, which is the most famous portal site in Korea. When we try to find a word, a restaurant, or directions to a place we don’t know, we can access Naver to get the information we want. To find out about the Naver location recommendation algorithm, one of the most frequently used functions by users, I met Lee Jun-Geol who works at Naver. After entering the Department of Statics at Gachon University in 2013, he has achieved a variety of specifications, including presidential activities in the Big Data Association, scholarship activities of the Korean Leader Fostering Scholarship Association, and academic festival promotion activities. More recently, he also graduated from Korea University graduate school. Currently, he is working for the Naver Search CLC AirSPACE team, which develops and manages Naver location recommendation algorithms. Through this interview, let’s look for the information on Naver’s place recommendation algorithm that we might not otherwise know well, various activities he did, and the advice he gave.
1. You have done a lot of activities so far. Which was the most memorable?
I have done a lot of activities, but among them, the most memorable one is working for the Korean Leader Development Scholarship Foundation. The Korean Leaders Development Scholarship Foundation selects students who are active in school or have good grades on a provincial basis and there are various benefits such as receiving scholarships if you enter. Networking with students from all other schools two or three times in a year was memorable and helpful. If you work for the Korean Leader Development Scholarship, you can receive scholarships to grad school, so it would be better for students who want to go to grad school.
2. What was the most useful activity that you did?
I think the meeting called ToBig’s(투빅스) that studies big data, data analysis, and AI was the most helpful. In this meeting system, it was good for seniors to be a tutor of their juniors for 15 weeks. We did various activities such as feedback, assignments, and mentoring. Through this, we were able to experience development outsourcing within the society or experience a lot of developments. As an external developer of SK Networks, I also had an opportunity to develop an app to apprehend my dog’s emotional state by taking pictures. People with similar interests gathered, we talked a lot about the future, recommended grad school to each other, and went out to lectures together. I think it was the most helpful because my skills were improved through these activities. I recommend it to seniors who are preparing for employment because it helps to do high-quality things for filling out resumes.
3. What made you decide on your career path in AI fields?
It was the second semester of the second grade when I decided on my career path in AI fields. As I studied statistics like my other colleagues, I wanted to have my own unique skill to stay competitive. When I thought about which strengths I could have in statistics dealing with data, I could find it would be better to study programming. Finally, I focused on studying AI.
4. What exactly is a place recommendation algorithm using AI?
Simply, it is the same as a YouTube recommendation algorithm. When we first enter YouTube, we can watch some videos that suit us. Like this, places that suit individuals are searched. However, more complex concepts are added than YouTube. YouTube can be seen anywhere we want, but the place recommendation algorithm, is limited compared to the video algorithm’s role because the function of ‘distance’ is added. If we search for a famous restaurant, this system’s role is to recommend restaurants closest to our house, so it needs relatively more systems than a video platform. In short, it can be said that providing a thoroughly personalized service is a place recommendation algorithm.
5. What do you think is the most important in Machine Learning and Deep Learning technology?
I think I can answer this question for students and developers respectively. First, if you are a student, you just need to make a good model. It does not matter how long it takes, and it is great to just utilize the given data to make a model. The second is when you get a job as a developer. Developers need to create models that can be used directly in reality. I think it is important in Machine Learning and Deep Learning technology to be able to get feedback from users quickly and to create a model that responds immediately when executed.
6. Is there a special reason that you set Naver as a goal?
Naver is a famous site that everyone knows, and it is liberal. Also, it has an advantage of providing it to everyone anytime, everywhere. So, I selected Naver as a goal.
7. What are strengths and weakness in your work?
The strength is that I can easily explain what I am doing. If people are curious about what I do, I think the biggest advantage is that I can easily make people understand Naver service which is widely used through mobile phones. On the other hand, the weakness is that as a developer, there is a lot of overtime work. Also, Naver’s location recommendation algorithm is a 24-hour system, so if we have a problem, we must carry a laptop to work regardless of weekends or overtime.
8. If you can hold an academic festival again, what topic would you like to do?
If I participated in an academic festival again, I would like to talk about dealing with data statistically. It usually deals with data using numbers, but I want to convey that it is available to convert photos or texts called ‘unstructured data’ into general data covered in statistics.
9. How did you overcome hardship when preparing for employment?
I think writing a cover letter was the hardest process while preparing to get a job. There are generally five questions in cover letters, and it was stressful to think about how to fill them. After considerations, I found my own solution which is to create the structure of a cover letter. When I structured what activities to write for each question, I was able to get the greatest effect. I could finish my self-introduction easily by applying one activity to one question. For example, applying external activities to the first question, an activity at school to the second question, and the experience participating in the contest to the third question.
10. Please say a word to students who want to get a job in the same field.
I'd like to answer this question in two cases: students with the lower grade and the upper grade. In the case of lower grades, it is good to study major or development, but I think it would be better to experience various activities. As a developer, I would like to recommend challenging various activities rather than vaguely studying because individual abilities are all different even though they are the same developer, such as those who collaborate well, speak well, and have excellent leadership. If you are a senior, I hope you can submit a resume and get used to failure after applying to the company. I hope you think that I am not a problem, the company and I are not meant to be, and that there will be another company that will connect with me. Rather than being afraid of failure, I hope you try hard until you find a company that suits you.
Through this interview, we were able to get useful answers from a senior, Lee Jun-Gul. Various kinds of questions such as Naver place recommendation algorithm and external activities, seem to be a good answer for Gachon university students with numerous concerns about finding a career and activities to choose. Every word from our senior was useful and good, but among them, I remember the most is his recommendation to experience various activities. Realizing through various activities and developing your own abilities in the future will play an important role anywhere. We welcome you students to be able to form your own weapons through diverse experiences of your own.< Copyright © The Gachon Herald All rights reserved >