This CSV dataset is designed to train similarity models using pairs of court decision summaries. The dataset includes similar legal cases with comparable contextual frameworks and principled legal solutions derived from consistent judicial standards. The cases are categorized by the following roles: