부산대학교 도서관

With the continuous development of digitization, there are many invisible threats behind the increasingly developed and complex network. Network security texts are mainly published in the form of Network Threat Intelligence on major network community platforms. In this paper, information extraction technology is used to automatically identify security related entities, relationships, attributes and other information from the network security text. Unstructured text data is transformed into structured expression which is convenient for storage and analysis, and presented in the form of graph data. It provides guidance for technicians to save time, and provides network security decision support for high-level.