摘要
公安询问记录是案件侦破的关键情报载体,能帮助公安机关快速掌握事件现状和成因,预判发展趋势,为决策和行动提供数据支持。结合外部常识知识对询问记录进行情感摘要生成,可提升决策效率与准确性。提出一种基于外部常识知识融合的情感摘要生成方法,应用于公安询问记录,通过融合常识知识,有效拓展语义范围,设计动态图注意力模型,使摘要与常识知识高度契合。实验表明,该模型显著提升了摘要的准确性和实用性。
The interrogation records of public security serve as a crucial carrier of intelligence for criminal detection.They can help public security organs to quickly grasp the current state and causes of an event,predict the development trend of the event,and thereby provide strong data support for public security decision-making and taking actions.Integrating external common sense knowledge to generate emotional abstracts of interrogation records is conducive to improving the efficiency and accuracy of decision-making.A common-sense knowledge fusion-based emotional abstracts generation method is proposed in the study,taking public security interrogation records as the application scenario.By fusing a common sense knowledge,the semantic scope of the interrogation is effectively expanded.Furthermore,a dynamic graph attention model is designed,the generated abstracts are highly consistent with common sense knowledge.The experimental results show that the model can significantly improve the accuracy and practicality of abstracts.
作者
尉译心
刘三满
李宁
WEI Yixin;LIU Sanman;LI Ning(Shanxi Police College,Taiyuan 030401,China)
出处
《火力与指挥控制》
北大核心
2025年第9期157-163,169,共8页
Fire Control & Command Control
基金
辽宁网络安全执法协同创新中心基金资助项目(XTZX2024-001)。
关键词
常识知识融合
询问记录
情感摘要生成
图注意力模型
智慧警务
commonsense knowledge integration
interrogation records
emotional abstracts generation
graph attention model
intelligent police
