| 摘 要: 随着国家司法信息化建设的推进，促进司法公平正义、提高司法审判工作效率成为必然的趋势。本文将事件图谱的构建方法引入司法领域，将裁判文书案情描述中的犯罪行为视为节点，犯罪行为间的逻辑关系视为边，构建了一种犯罪行为演化图谱，以此来将案情中孤立的犯罪行为利用逻辑关系连接起来，从而描述案情的发展过程而不是独立的要素，最大限度地保留了案情的语义信息。本文提出的自动构建方法在五个案由中的抽取结果比传统方法平均提高了7.8%，此方法构建的犯罪行为演化图谱可为法律从业人员研究案情提供参考，同时为推进智慧法院建设如罪名预测、类案推荐等提供了新的思路与方法。 |
| 关键词: 犯罪行为；关系识别；犯罪行为链；犯罪行为演化图谱 |
|
中图分类号: TP399
文献标识码: A
|
| 基金项目: 国家自然科学基金资助项目(62066008). |
|
| Research on Automatic Construction Method of Criminal Act Evolution Graph |
|
DUAN Xihui, HUANG Ruizhang
|
(School of Computer Science and Technology, Guizhou University, Guiyang 550025, China)
1416831033@qq.com; rzhuang@gzu.edu.cn
|
| Abstract: With the advancement of national judicial information construction, it has become an inevitable trend to promote judicial fairness and justice, and improve the efficiency of judicial trials. This paper proposes to introduce construction method of event graph into the judicial field and construct an evolution graph of criminal acts, where criminal act in case fact description of the judgment document is regarded as a node, and the logical relationship between criminal acts as an edge. Isolated criminal acts are thus connected by logical relationship, so as to describe the development process of the case rather than independent elements, and preserve the semantic information of the case to the greatest extent. The extraction results of the five reasons of case by the proposed automatic construction method are improved by an average of 7.8% compared with the traditional method. The criminal act evolution graph constructed by this method can provide a reference for legal practitioners to study cases, and at the same time, it provides new ideas and methods for advancing the construction of smart courts, such as crime prediction and case recommendation. |
| Keywords: criminal act; relationship identification; criminal act chain; criminal act evolution graph |
