UniMoral is a multilingual dataset designed to study moral reasoning as a computational pipeline. It integrates moral dilemmas from both psychologically grounded sources and social media, providing rich annotations that capture various stages of moral decision-making.
Psychologically grounded dilemmas in UniMoral are derived from established moral psychology theories, including Kohlberg’s Moral Judgment Interview (MJI)… See the full description on the dataset page:
https://huggingface.co/datasets/shivaniku/UniMoral.
