融合知识图谱的网络安全违法行为识别系统

《网络安全法》是我国第一部关于网络安全的综合立法,与大众的互联网生活息息相关.因此,一款面向大众的《网络安全法》智能违法行为识别系统有助于规范互联网行为.然而,现有智能违法行为识别系统构建方法难以适应《网络安全法》,这是由于:首先,现有方法需要专业司法语言进行交互,不适应普通大众的语言体系.其次,现有方法需要大量的案例训练模型,不适应案例匮乏的《网络安全法》.针对这些问题,本文提出了一个面向《网络安全法》的智能违法行为识别系统.该系统主要利用知识图谱技术解决上述问题,在构建网络安全法知识图谱的基础上,通过将普通用户的自然语言与知识图谱中的违法事件实体和违法主体实体进行实体链接的方式获得更强的特征,提高违法行为识别系统在训练集较为匮乏的条件下的准确度.通过在真实数据集上的实验,表明了提出的系统的准确度有明显提高.

Cyber Security Illegal Behavior Recognition System Integrating Knowledge Graph

The Cybersecurity Law of the People’s Republic of China is the first comprehensive legislation on cyber security of China,which is closely related to the Internet life of the public.Therefore,an intelligent illegal behavior recognition system for the Cyber Security Law helps public to regulate their Internet behaviors.However,it is difficult to adapt the existing intelligent illegal behavior recognition systems to the Cyber Security Law due to the following two reasons.First,the existing systems require professional judicial language for interaction,so it is not suitable for colloquial language of the public.Second,the existing systems require a large number of samples for model training,while the Cyber Security Law is a new law and lacks training samples.To this end,this paper proposes an intelligent illegal behavior recognition system specifically designed for the Cyber Security Law.It is designed based on knowledge graph technique.It solves the above mentioned problems by linking the colloquial language of the public to the illegal event entities and illegal subject entities in the knowledge graph.The experiments based on real dataset have demonstrated the effectiveness of the proposed system.