WCLD
收藏arXiv2023-10-28 更新2024-06-21 收录
下载链接:
https://clezdata.github.io/wcld/
下载链接
链接失效反馈官方服务:
资源简介:
WCLD数据集是由苏黎世联邦理工学院法律与经济学中心创建的大型刑事案件数据集,涵盖了美国威斯康星州巡回法院自1970年至2020年的150万条案件记录。数据集详细记录了包括犯罪前科次数、再犯结果、性别、年龄(审判时及首次犯罪时)等信息,并提供了种族、社区特征、犯罪类型、指控严重性、案件决定、判决长度等多样化的属性。此外,数据集还包括法官、县和邮编的伪标识符,以及从人口普查数据中提取的邻里特征。WCLD数据集不仅支持对刑事司法背景下算法公平性的严格研究,还能关联算法挑战与各种系统性问题,为学术研究提供了丰富的数据资源。
The WCLD dataset is a large-scale criminal case dataset developed by the Center for Law and Economics at ETH Zurich, covering over 1.5 million case records from the Wisconsin Circuit Courts in the United States spanning from 1970 to 2020. The dataset comprehensively records a diverse set of attributes including prior criminal offense counts, recidivism outcomes, gender, age at the time of trial and at first offense, as well as race, community characteristics, crime types, charge severity, case dispositions, and sentence lengths. Additionally, the dataset contains pseudonymous identifiers for judges, counties, and zip codes, as well as neighborhood-level features extracted from census data. The WCLD dataset not only enables rigorous research on algorithmic fairness in the context of criminal justice, but also allows for linking algorithmic challenges to various systemic issues, serving as a rich data resource for academic research.
提供机构:
苏黎世联邦理工学院法律与经济学中心
创建时间:
2023-10-28



