| 摘要: |
| 恐怖事件、突发事件、冲突事件等特定主题事件通常对国家安全带来严重威胁，记录现实事件的事件库在态势感知、风险预警、应急决策等应用中发挥重要作用，事件库构建技术随之发展为内容安全技术的重要组成部分。事件库构建技术是一类实现从海量的非结构数据批量生成结构化事件数据的技术，由于数据环境、表示精度、应用场景的差异，出现了各类构建技术的相关研究。本文详细介绍了事件库的定义、分类和架构，按自底向上输出的数据层次，将事件库构建技术划分为事件检测、事件抽取、事件融合三类关键技术，并分别对其研究现状和进展进行了全面分析，总结了事件库的主要应用领域，最后对事件库构建技术中面临的主要挑战和关键问题进行了探讨。 |
| 关键词: 事件库 事件检测 事件抽取 事件融合 |
| DOI：10.19363/J.cnki.cn10-1380/tn.2019.03.08 |
| Received:July 11, 2017Revised:October 30, 2017 |
| 基金项目:国家自然科学基金（No.U163620068），战略合作专项（No.AQ-1704，AQ-1708）项目资助。 |
|
| Event Database Construction Techniques |
| XUE Cong,GAO Neng,ZHA Daren,WANG Lei,YIN Zhiyi,ZENG Zehua |
| State Key Laboratory of Information Security, Institute of Information Engineering, Chinese Academy of Sciences, Beijing 100093, China;Data Assurance and Communication Security Research Center, Chinese Academy of Sciences, Beijing 100093, China;School of Cyber Security, University of Chinese Academy of Sciences, Beijing 100049, China |
| Abstract: |
| Specific topic events,such as terrorist attacks,emergencies and conflicts,pose serious threats to national security.Since event database which records real-life activities plays a major role in situational awareness,crisis early warning and contingency decision,its construction techniques have become an important part of content security.The construction techniques of event database achieve mass production of structured event data from massive unstructured data.Because of the differences in data landscape,data precision and application scenario,there have been relevant researches of various construction techniques.This paper describes the definition,classification and architecture of event database in detail.According to the bottom-up data level,the paper divides the construction techniques of event database into three classes of critical technologies,including event detection,event extraction and event fusion,and makes comprehensive analysis on current research status and development of these technologies.Furthermore,the paper summarizes the main application fields of event database.Finally,this work conducts in-depth discussions on major challenges and key issues which the construction techniques of event database confront. |
| Key words: event database event detection event extraction event fusion |
