I turn qualitative customer data into psychological insight and strategic decisions.

Moral Foundations signals in customer reactions over time
Aggregate customer reaction structure from a real case
Chart summary and data

All four signals fell after the first May 18 interval. In most later three-hour buckets, fairness remained highest at roughly 0.46-0.54, followed by betrayal, degradation, and harm. Values are three-hour means of model outputs on a 0-1 scale.

Selected work

Probato

An independent project started in 2024. It separates meaning, stance, and moral foundation signals in short contextual reactions, then turns them into decision evidence.

Case study
Customer Moral Foundations profiles by semantic cluster
Frame numbers are unsupervised cluster IDs, not topic names or ranks.
Chart summary and data

Frame 7 is the dominant cluster with 57,128 items. Its mean signals are harm 0.22, fairness 0.52, betrayal 0.47, authority 0.33, degradation 0.39, and liberty 0.18. No topic meaning is inferred from the numeric ID.

Comparison of high-intensity stakeholder reactions by platform
Platform shares of reactions scoring 3.0 or higher
Chart summary and data

In Naver News comments, 46.8% showed high fairness criticism, 41.2% high governance criticism, and 20.3% high exit intention. In Clien comments, the corresponding shares were 43.7%, 35.7%, and 28.8%. Different samples and expression styles rule out a simple platform ranking.

SNU MBA Applied Business Project

An MBA graduation project that combines market reactions and online stakeholder data to compare cybersecurity crisis response alternatives.

Case study

From qualitative signals to strategic judgment

Qualitative analysisI separate irony, ridicule, stance, and context into useful units of analysis.
Customer psychologyI interpret the psychological frames behind aversion and accountability demands, not sentiment direction alone.
Data and AII connect collection, schemas, LLM-assisted labeling, embeddings, and evaluation in a reproducible workflow.
StrategyI translate findings into alternatives, priorities, execution conditions, and limitations.

Analysis with an implementation path

My background spans ethics education, editorial ownership, MBA study, and data analysis. I use it to explain why customers react and what the reaction means for a decision.