DHB: A Computational Approach to Understanding the Dynamics of the Judicial System

以前对司法系统的研究面临着一种取舍，一种是专注于容易统计的行为的大规模量化调查，另一种是允许更仔细地审查法律文本的较小规模的研究。该项目标志着首次尝试将信息检索和计算语言学应用于美国最高法院的研究。此外，将通过汇编最高法院文本集(SCTC)来促进未来的研究。在这个项目的过程中，调查人员将汇编、组织和注释与美国最高法院过去半个世纪审理的案件有关的15,000多份法律文件，并开发一套半自动分析工具来更仔细地审查这些文件。SCTC将向更大的学术社区提供更多的研究努力。应用计算技术对美国司法系统进行建模，为克服政治学中传统手工、劳动密集型方法的许多瓶颈提供了机会，也为信息检索和计算语言技术的进步提供了新的环境。这个项目将采用一种新的基于文本的法律系统计算模型，具有解释和预测能力，使我们能够提出广泛的理论上有趣的研究问题。更重要的是，这种将计算技术应用于法律制度研究的跨学科方法，有可能彻底改变未来社会科学家的研究议程和教育。为了促进这一进展，研究人员计划开发一门课程，名为“政治文本分析的计算方法”，以及相关的研究研讨会，将为其他学者、研究生和精选的本科生提供必要的技能，将计算方法应用于他们自己的研究问题。

Previous research of judicial systems has faced a trade-off between large scale quantitative inquiries focused on readily-counted behaviors, and smaller studies that allow closer examination of legal texts. This project marks the first attempt to apply information retrieval and computational linguistics to the study of the US Supreme Court. In addition, future research will be facilitated by assembling a Supreme Court Text Collection (SCTC). During the course of this project, the investigators will compile, organize, and annotate more than 15,000 legal documents associated with cases heard by the US Supreme Court over the last half-century, and develop a set of semi-automated analytical tools to examine them more closely. The SCTC will be available to the larger academic community for additional research endeavors. The application of computational techniques to model the US judicial system represents an opportunity to overcome many of the bottlenecks associated with traditional manual, labor-intensive methods in political science, and also provides a new environment for the advancement of information retrieval and computational linguistic techniques. This project will employ a novel text-based computational model of the legal system with both explanatory and predictive power that will allow us to pose broad classes of theoretically interesting research questions. More importantly, this interdisciplinary approach, applying computational techniques to the study of legal systems, has the potential to revolutionize both the research agendas and the education of future social scientists. To facilitate this progress, the investigators plan to develop a course, "Computational Approaches to Analysis of Political Texts," and related research workshops that will equip other scholars, graduate students, and select undergraduate students with the skills necessary to apply computational methodologies to their own research questions.