付琳1, 2 王俊3*(1. 中南财经政法大学，武汉 430000;2. 贵州警察学院，贵阳 550000;3. 西南政法大学刑事侦查学院，重庆 401120)

【摘 要】随着大数据技术的出现及司法应用，基于大数据进行案件事实的认定，已成为事实认定者 不可回避的话题，特别是刑事诉讼领域。然而，因大数据技术的司法应用，案件事实的认定方式发生了 明显变化。而且，基于大数据的刑事案件事实认定，面临着认定场域全景化影响庭审和事实认定，认定 过程隐秘化易导致算法黑箱、数据独裁和算法霸权，事实认定形式化及数据表达不兼容影响大数据功能 的实现，相关性分析结果存在偏差和失灵风险，事实认定面临全量样本难以获得和证明的系统性风险等 问题。鉴于此，有必要重新解构刑事案件事实认定涉及的“犯罪重建理论”“案件结构理论”以及“证据 之镜理论”，并考虑从主体维度、过程维度、应用维度和逻辑维度予以改进，坚持以人为本，促进可视正 义和大数据分析报告的司法应用，以相关促因果，进而保障刑事案件事实认定的科学性和准确性。

【关键词】大数据;刑事案件;事实认定;原理;问题;多维思考

【中图分类号】D915.13

【文献标识码】A

【文章编号】1674-1226(2024)02-0229-14

Multidimensional reflections on using big data analytics for criminal trial factfinding. Fu Lin1,2, Wang Jun3. 1. Zhongnan University of Economics and Law,Wuhan 430000, China; 2.Guizhou Police College, Guiyang 550000, China; 3.School of Criminal Investigation, Southwest University of Political Science and Law, Chongqing, 401120, China.

【Abstract】With the emergence of big data technology and its application in judicial proceedings, using big data analytics for factfinding has become an unavoidable topic for fact-finders, especially in the field of criminal proceedings. However, due to the application of big data technology in judicial proceedings, the way of factfinding has changed significantly. Moreover, the factfinding based on big data analytics would put criminal trials at risk. For example, the panoramic application of big data analytics would affect the trial and factfinding; The secret application of big data analytics would easily lead to algorithm black box, data dictatorship and algorithm hegemony; The formality of factfinding and incompatible data expression would affect the realization of big data functions; The relevancy analysis would face the risk of bias and failure; The factfinding would face the systemic risk in the difficulty in obtaining and proving full samples. In view of such situation, it is necessary to re-deconstruct the “crime reconstruction theory”, “case structure theory” and “mirror of evidence theory” involved in criminal trial factfinding, and improve these theories from the dimensions of the subject, the process, the application and the logic. As adhering to the people-oriented approach, the effectuation of visible justice and the application of big data analytics reports in judicial proceedings would be promoted, and the causation analysis based on relevance would be established. Accordingly, the scientific and accurate criminal trial factfinding can be guaranteed.

【Key Words】Big data; Criminal cases; Factfinding; Principle; Problem; Multidimensional reflections