NC

Our team discovers insights through analyzing game data. We develop indicators and services based on machine learning and statistical modeling.

I design the overall flow of data analysis. I develop the entire process,

covering data extraction, statistical modeling, and service development based on analysis results. Nowadays, I mainly carry out analyses based on deep learning.

Games resemble the reality. In the world inside games, economic activities naturally take place, as users hunt and carry out transactions. As a result, the phenomena that appear in the real world, including inflation and deflation, occur. Also, group of friends are formed through social activities, and sometimes, leaders emerge. As time goes by, the way individual players play games may change, and some users quit playing the game with others.

All these are accumulated in the form of data. If you take a look at the data, the world inside games are like a condensed version of the real world. That is why game data are interesting. Sometimes, the things that cannot be experimented in the real world, such as “how would people behave in case the world ends?”, are actually analyzed based on game data.

Every matter that occurs in a game will be recorded in the form of data. However, they will remain as ‘dots’ unless we discover problems in them. Our team discover problems based on data. We either model the users’ behavioral patterns or analyze the effect from an event based on passage of time. We actually discover bots in games and the users that are likely to quit. We carry out in-depth analyses of promotional effects.

In the beginning, I carried out basic tasks, such as indicator development and data pre-processing. I have continued to learn things one by one and expanded my scope of work into in-depth analyses. I carry out user clustering, network analysis, and apply deep learning to systems. We have filed a patent application for application of deep learning, under the title of .

Developers that only know about coding are difficult to work with. To understand the words of analysts, developers should have statistical knowledge; to talk with game operation departments, they should understand games.
Our team study together to enhance capability of each team member. I ran a study group with the team members when we tried to apply deep learning. We always study to try new methodologies.

As an undergraduate, I had interest in the field of data. I took a class in statistics and prepared for the DBA. Around the time I graduated from school, the data industry started to attract attention. Therefore, there were many conferences and seminars to attend. I attended many of them and developed interest in data analysis. Even people like me that majored in computer science could have a shot.

Around that time, there was a recruitment session hosted by NC, and the program offered a Q&A session with incumbent employees of each function. I asked many questions, as I was interested. In three months, I participated in an interview. One of the interviewers was the one that I talked to during the recruitment session. Fortunately, the interviewer remembered me. I do not know whether my active approach worked or not, but anyway, I joined the company in 2015.

There are many occasions where we cannot get what we expect from research. In those cases, we think that what we are trying would work, but they don’t. It seemed that the time we invested in the project would be in vain. We should not only carry out research, but also produce service based on its result. Therefore, we sometimes feel hurried. Of course, it still is difficult. Nowadays, whenever I am stuck somewhere, I talk to my team members or study to address the problems.

By the time I graduated from college, data analysis and deep learning were not so common. However, they are now easily accessible even by non-professionals. Those that join the company these days have splendid backgrounds, as they have won contest awards, etc.

Like we learn English and math in high schools, someday, we would learn data analysis and deep learning as basic knowledge. Therefore, if you are a developer that has a ‘shallow’ knowledge, you would end up as a developer that has no merits to work with.

I don’t want to be such a developer. That is why I keep studying and applying new techniques. I also make efforts to explain those things as easily as possible.

I travel to new destinations every year and try gourmet foods in fine dining restaurants with friends. I have tried many whiskeys and wines. A few years ago, I attended every single music festival that took place in Korea. I appeared in a movie directed by my friend and painted pictures in an art studio.

I did not do those things because I had a specific target in mind. The more I know about something, the more it gets interesting. It is also fun to get to know about things. Whenever there is an opportunity for new experience, I try not to avoid it.

I want to be a person that is good at expressing his knowledge either in a written or a verbal form. I dream of myself making confident arguments based on logical evidence. When you are able to express your opinion about a certain area, it means that you are an expert. I want to reach the deepest level of knowledge.

However, I do not want to try too hard. I just live every day to the fullest. By doing so, I would become a person that I dream of. It does not matter if I don’t. However, I would continue to live my life to the fullest, continuing to have the attitude to expand my knowledge even further in my field of expertise.