International Conerence Big Data Analytis, Data Mining and Computational Intelligence, 233-235

" This was my very first conference paper written during my time in industry. While working on this project, I realized how much I enjoyed the research process—so much so that it inspired me to pursue a Ph.D. It’s a small piece of work, but one of the most memorable papers in my journey. "

We analyze hotel reviews to help consumers make better decisions by scoring the sentiment of keywords. Using NLP and regression, we link keywords to review ratings and assign each one a Sentiment Score. For keywords with weak signals, we use a k-NN–based method to estimate their sentiment. The study is based on hotel review data from South Korea.

The Korean Journal of Applied Statistics, 33(1),75-86

" This paper was an extension of my earlier work, developed after I entered graduate school and further analyzed in collaboration with a colleague from my previous company. While I now see areas where it could be improved, this paper still reminds me of the passion and excitement I felt at the time. "

We use sentiment keywords from hotel reviews to build a summary view of what customers like or dislike about hotels. To help users avoid reading tons of reviews, we create a hotel network based on shared sentiment patterns using topological data analysis. This lets users explore hotels with similar pros and cons, and gives marketing teams insights into how their hotel is perceived.

IEEE Access 

We explore how employee-related factors are linked to company revenue using machine learning. By testing models like XGBoost on both annual and integrated data, we classify company performance and identify key HR-related features. The results highlight the strong predictive power of our approach and offer practical insights for strategic human resource management.