Showcasing how data science connects research with real-world policy.

On September 12, 2025, KDI School of Public Policy and Management hosted the “Data Science Poster Session”, bringing together students, faculty, and guests to celebrate research that applies cutting-edge data methods to real-world policy challenges. The three-hour showcase featured posters from KDI School and partner universities, including Kangwon National University, Korean Advanced Institute of Science and Technology (KAIST), and Yonsei University, covering topics as diverse as gender and political leadership, railway segmentation, birth rate trends during COVID-19, and the evolution of North Korean policy discourse.

The event was more than a research presentation, it was an intellectual exchange that demonstrated how data science continues to shape public policy thinking in Korea and beyond.

Exploring the Range of Research

The posters reflected both methodological innovation and pressing social issues. Among KDI School participants, research included “How Digital, AI, and GenAI Divides Shape Technology Adoption,” which examined equity in emerging technologies, and “Predicting User Engagement and Popularity on News Articles Using Metadata and Semantic Embeddings,” which harnessed BERT models to anticipate trends in media consumption.

Other projects tackled geopolitics and governance. For example, Songhyeon Kim (MDS, 2025) presented “Tracing North Korean Policy Discourse Change under the Kim Jong-un Era.” Despite limited access to structured open data, her work used large-scale text analysis to uncover shifts in the Democratic People’s Republic of Korea’s domestic policy rhetoric. As Kim explained, “Big data allows us to observe policy trends in a quantitative way that would be impossible for individuals to read in full.”

In the transport sector, Hyunju Lee (MPP, 2025) presented “A Study on Segmentation of Railway Customers Using Pseudonymized Information Combination and Clustering Techniques.” By combining internal datasets with pseudonymized external information, Lee was able to reveal distinct lifestyle-based clusters of passengers, such as leisure versus business travelers, that traditional statistics overlooked. She emphasized how these insights could inform differentiated fare policies, scheduling, and equity-driven transport planning. “Data science will increasingly allow policymakers to move beyond description toward predictive and prescriptive analytics,” she noted.

One team of MDS (2025) students, Ilva Prifti, Alieu Ceesay, Brigid Onyekachi Egbikwe, and Veronica Kiteve, focused on political leadership in Africa with their study “Predicting Gender Political Leader Preference in Africa.” Their findings challenged assumptions; despite the low number of female heads of state, data consistently showed that African citizens expressed equal preferences for male and female leaders. “The most significant predictors of preference were: individuals’ social beliefs regarding whether women should have access to jobs when employment is scarce, women’s rights to land ownership, as well as gender, country of residence, and education level. These findings indicate that societal gender beliefs strongly influence political leadership preferences,”the team explained. Their research highlighted the role of evidence-based advocacy for women in leadership.

Beyond KDI School, contributions included projects on AI complementarity between public and private sectors (Kangwon National University), semantic shifts in birth-related keywords during COVID-19 (Kangwon National University), and a computational history of democracy (KAIST/Yonsei University). 

Student and Attendee Perspectives

The enthusiasm of both presenters and attendees amplified the vibrancy of the session. David Ugarte (PhD-PP, 2024) reflected on the event’s value: “It demands research that is not only robust but also relevant. There’s an ocean of statistically significant yet irrelevant work out there, what we saw here bridges rigor with practical meaning.”

Chaeah Song (MDP, 2025) was drawn to the poster on predicting news engagement, praising its global applicability. More importantly, she emphasized the passion of the presenters: “They reminded me of the wide range of applications in data science, and how valuable these opportunities are for students, especially those in policy-related fields.”

On the other hand, Raquel Grünauer (MDP 2025) highlighted the session’s ability to merge abstract concepts with data-driven tools. She found inspiration in projects like “Who Are the People? A Computational History of Democracy” and the poster on technology divides. “It showed me that data science is not just about numbers but about understanding reality,” she said.

For many, the event also provided methodological takeaways. Ugarte, for example, noted how discussions around TF-IDF, clustering, and masking-based analysis would feed directly into his own ongoing research.

Looking Ahead

The 2025 Data Science Poster Session reaffirmed KDI School’s commitment to equipping students with the tools to bridge theory and practice. It also strengthened collaboration across institutions, with contributions from Korean universities demonstrating the interdisciplinary power of data science.

The session ended not with conclusions but with inspiration, students leaving with fresh ideas, methodological insights, and a renewed sense of how their work can influence public policy. The posters demonstrated that data science is not just about creating models, but about shaping decisions that impact people’s lives.

In a world where evidence-based policy has never been more critical, the 2025 KDIS Data Science Poster Session showcased what the next generation of scholars can contribute: rigor, creativity, and above all, relevance.