Online conversational data is more than language alone. It carries privacy, emotion, relationships, and the social context that shapes how people communicate.

dda collected, refined, and labeled large-scale Korean chat data from online messaging environments, and established standards for de-identification and ethical content handling. The result is a data foundation designed for safe public release and responsible, real-world use.

- Collecting large-scale online conversations grounded in everyday Korean messenger usage
- Establishing de-identification standards to protect personal information and private context
- Designing refinement and judgment criteria for handling aggressive or unethical expressions
- Designing labeling structures and review workflows to ensure consistent quality
- Establishing public data operations frameworks that support reuse and future expansion