摘要
数字赋能逮捕社会危险性评估有助于解决评估数据容量较小及质量不高、结果缺乏检验等问题。其逻辑在于,利用数字技术收集反映逮捕社会危险性的各项证据并转化成计算机可理解的结构化数据,再通过算法测算出风险等级并输出人类可理解的司法建议。从效果来看,数字赋能逮捕社会危险性评估在增强评估模型专业性、保障评估模型客观性和提高评估结果准确性等方面优势明显,但同时也面临着“僵尸预测”、算法歧视、算法黑盒等方面的质疑。未来改革需要进一步调整思路,强调在构建专业评估模型的基础上确保评估数据的数量与质量,并建立动态的校准机制,以不断优化评估的模型;同时,还应当强化评估算法的公开、算法歧视的治理、算法的审核与问责,完善评估算法的治理模式;此外,还需保障评估过程中犯罪嫌疑人的参与权、隐私权和救济权,确保评估程序的公正,以此实现数字赋能逮捕社会危险性评估的预期目标。
Digital empowerment in the assessment of social risk in arrests helps address issues such as limited data volume,low data quality,and lack of validation in assessment results.Its logic lies in utilizing digital technologies to collect evidence reflecting the social risk of arrests and transforming it into computer-understandable structured data.Algorithms are then employed to calculate risk levels and output judicial recommendations comprehensible to humans.In terms of effectiveness,digital empowerment in the assessment of social risk in arrests demonstrates significant advantages in enhancing the professionalism of assessment models,ensuring their objectivity,and improving the accuracy of assessment results.However,it also faces criticisms such as“zombie predictions”,algorithmic bias,and the“black box”nature of algorithms.The future reform needs to further adjust the thinking approach,emphasize ensuring the quantity and quality of assessment data on the basis of constructing professional assessment model,and establish a dynamic calibration mechanism to continuously optimize the assessment model.Additionally,efforts should be made to enhance the transparency of assessment algorithms,address algorithmic bias,and strengthen the review and accountability of algorithms to improve governance models.At the same time,we should also strengthen the transparency of assessment algorithms,the governance of algorithmic bias,the review and accountability of algorithm,and improve the governance model of evaluation algorithm.Furthermore,it is essential to safeguard the participation rights,privacy rights,and remedies of suspects during the assessment process,ensuring the fairness of the assessment procedures,thereby achieving the intended goals of digital empowerment in the assessment of social risk in arrests.In addition,it is necessary to safeguard the participation rights,privacy rights and relief rights of crime suspect in the assessment process,and ensure the fairness of the assessment procedure,so as to achieve the expected goal of digital empowerment in the assessment of social risk in arrests.
作者
徐暾
Xu Tun(Law School,China Jiliang University,Hangzhou,Zhejiang 310018)
出处
《河南财经政法大学学报》
2025年第2期40-54,共15页
Journal of Henan University of Economics and Law
基金
2023年国家社会科学基金项目“刑事诉讼中羁押审查的分级量化评估研究”(项目编号:23BFX172)的阶段性成果。
关键词
数字赋能
逮捕条件
社会危险性
算法
digital empowerment
conditions of arrest
social risk
algorithm
作者简介
徐暾,男,中国计量大学法学院副研究员、硕士生导师,研究方向为刑事诉讼法。
