| 摘要: |
| 国内现有的智能司法系统普遍采用数据型信息作为基础设施，由于其数据本身存在来源性、模糊性与孤立性的三重困境，致使应用系统难以实现实体性的法律决策辅助。鉴于人机思维的对应性特征，该问题可在法律获取理论的指引下得到解决： 智能司法系统的决策思维系发现与适用法律规范的过程，而作为其智识驱动来源的核心基础设施，也应由不确定的法律数据嬗变为确定的法律规则知识。在本体论层面上，原子化的法律构成要件是基础设施的最小知识单元，其互相以数理逻辑符号连接成前束范式的规则表示，且可通过生活经验与司法判决加以解释，以此形成体系化的要件知识网络。而在运行论层面上，该基础设施有赖立法、司法、执法与法学研究主体进行协作建设，其需要构建相互沟通的法律解释共同体，并依据法教义学方法形成法律规则的融贯体系，从而推动智能司法发展迈向知识范式。 |
| 关键词: 司法 人工智能 基础设施建设 法律获取 要件知识 法教义学 计算法学 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| From Data to Knowledge： Reflection on the Infrastructure Constructionof Intelligent Judicial Application |
|
Chen Liang & Xu Ming
|
|
|
| Abstract: |
| The current intelligent judicial applications in China generally use data based information as the infrastructure. Due to the triple dilemmas of source, fuzziness, and isolation in the data itself, it is difficult for the application system to achieve substantial legal decision making assistance. Based on the corresponding characteristics of human machine thinking, the theory of legal acquisition provides guidance for it. The process of decision making thinking of an intelligent judicial application is to discover and apply laws, and the infrastructure as the driving source of artificial intelligence should be transformed from the uncertain legal data to the certain knowledge of legal provisions. At the ontological level, the atomized constitutive requirement is the smallest knowledge unit, which is expressed by the rules of the first order predicate logic and can be explained by judgment to form a systematic knowledge network. At the operational level, the infrastructure depends on the collaborative construction of legislators, judges, administrators and scholars. It needs to build a community of legal interpretations that communicate with each other, and form a coherent system of legal rules based on the method of legal dogmatics, so as to promote smart judicial development moves towards the “knowledge paradigm”. |
| Key words: Intelligent Judicial Application, Infrastructure Construction, Legal Acquisition, Requirement Knowledge, Legal Dogmatics, Computation Legal Studies |
