| 摘 要: 伴随数字技术推动虚实社会深度融合,社区矫正对象再犯风险呈现隐蔽、动态、跨域特征,传统静态指标与一次性评估模式存在识别滞后、预警不足等问题,难以满足“事前预防”需求。本研究基于“刺激—机体—反应”(S-O-R)模型,将再犯视为动态交互过程,通过构建再犯风险事件库与推理双智能体系统,提取风险事件、梳理因果关系,结合图网络分析挖掘关键节点与核心路径,构建事理图谱及“人智协同”分析框架。研究结果显示,当前再犯风险评估存在明显的“时代性盲区”,而事理图谱能够有效串联碎片化信息、推演风险因果关系,清晰揭示数字时代再犯风险的动态演化逻辑,为社区矫正工作从被动监管向主动预警、从静态防控向动态治理转型提供了关键技术工具,助力实现再犯风险的早期识别与精准干预,推动社区矫正领域公共安全治理的升级。 |
| 关键词: 事理图谱 新行为主义理论 社区矫正对象 再犯风险事件 智能体 |
|
中图分类号: D926
文献标识码:
|
| 基金项目: 本文系“浙江省科技厅重点研发计划项目”“社区矫正对象重新犯罪精准靶向矫治关键技术与装备”(项目编号:2024C03267)的研究成果之一。 |
|
| Event Logic Graph-Based Research on Recidivism Risk Evolution Paths of Community Correction Subjects |
|
wangshixiong yangshuo sunpeiliang
|
1.Zhejiang Sci-Tech University;2.Zhejiang Police Vocational Academy
|
| Abstract: With digital technologies driving the deep integration of the physical and virtual societies, the recidivism risks of community correction objects have exhibited the characteristics of being hidden, dynamic and cross-regional. Traditional static indicators and one-off assessment models suffer from problems such as delayed identification and insufficient early warning, making it difficult to meet the demand for "pre-event prevention". Based on the Stimulus-Organism-Response (S-O-R) Theory, this study regards recidivism as a dynamic interactive process. By constructing a recidivism risk event database and a reasoning dual-agent system, it extracts risk events, sorts out causal relationships, explores key nodes and core paths with graph network analysis, and further builds an event logic graph and a "human-AI collaborative" analysis framework. The results show that current recidivism risk assessment has obvious "era-specific blind spots". In contrast, the event logic graph can effectively connect fragmented information, deduce risk causality, and clearly reveal the dynamic evolution logic of recidivism risks in the digital age. This study provides a key technical tool for transforming community correction work from passive supervision to active early warning and from static prevention and control to dynamic governance, facilitating the early identification and precise intervention of recidivism risks and promoting the upgrading of public security governance in the field of community correction. |
| Keywords: Event Logic Graph SOR Theory Community Correction Subjects Recidivism Risk Event Agent |
