Korean DUI Legal Judgment Prediction: Classical ML, LLMs, and Explanation Agreement (reproduction materials)
Reproduction materials for a study of Legal Judgment Prediction (LJP) on Korean drunk-driving cases: an attorney-verified benchmark of 1,500 cases drawn from the LBox Open corpus, eight classical machine-learning configurations, four large-language-model configurations, and a comparison of the two model families on both sentence prediction and explanation. The archive contains the dataset in open formats, the analysis code as it was run, the model predictions and cached LLM replies, the statistical result tables, and the figures, tables, and supplement. Every reported number can be regenerated from these files. The dataset in 1_dataset/ is also deposited separately, under https://doi.org/10.5281/zenodo.21647662, so that it can be cited and reused independently of these experiments.