情报分析对于国家安全具有重大的意义。事件作为一种重要的情报数据，比一般的实体知识更具有表达性，也更为复杂。事件图谱以事件为基本单位，包含事件、事件属性、事件关联关系等，能够实现高效的信息聚合，为事件的智能化处理和情报分析提供支撑。互联网新闻文本是一种重要的数据源，由于这些数据体量庞大，更新频繁，由人工收集和处理这些文本并分析其中的事件情报信息显得尤为困难，因此需要一种解决方案来实现自动化和智能化处理。本文提出了一种面向互联网新闻文本的事件图谱构建框架，在其基础上提出了面向情报分析的事件图谱应用方法。工作主要如下：（1）面向互联网新闻文本的事件图谱构建的系统框架与方法研究。提出了一种面向互联网新闻文本的事件图谱构建流程框架。该框架将流程分为四个基本步骤：新闻结构化信息抓取、新闻文本聚类、事件抽取和事件关系构建，使用统一的体系综合了各个步骤中的概念与方法，实现了事件图谱构建的规范化。（2）事件图谱在情报分析中的应用方法研究。在事件图谱的基础上，提出了实体-事件关联分析、事件脉络生成、事件关联分析方法及其在情报分析中的应用方法。（3）事件图谱构建及其应用的软件系统开发。设计了事件图谱构建系统的架构，实现了一套实际的事件图谱构建和应用系统，验证了以上方法的可行性与实用性。

Intelligence analysis plays a significant role in national security. Event is one important type of intelligence data, which is more expressive and complex than normal knowledge of entity. Taking an event as a basic unit, an event graph contains events, event attributes and event relationships, which aggregates information efficiently and provides powerful support for intelligent processing on events and intelligence analysis.News texts on the Internet are an important data source. Due to its huge amount and frequent updates, it becomes difficult for people to manually collect and process these texts and make an analysis of intrinsic intelligence information. Therefore, it is of necessity to develop a solution to achieve automation and intelligence.This thesis proposes a web-news-oriented method of building event graph and an application method for intelligence analysis based on the event graph. The main work of this thesis can be listed as follows:(1) Research on a system framework and methods of building event graph oriented news texts on the Internet.This thesis proposes a framework for building event graph oriented news texts on the Internet. The framework divides the procedure into four steps: structured news information crawling, news text clustering, event extraction and event relationship building, and normalizes the procedure by uniting various concepts and methods of the four steps into an intergrated system. (2) Research on application methods of intelligence analysis based on the event graph.Based on the event graph, this thesis introduces the methods of entity-event association analysis, event vain generation and event association analysis and how to apply them to intelligence analysis.(3) Development of an actual software system of event graph construction and application.This thesis designs an architecture of event graph construction and application system. An actual software system of event graph construction and application is implemented, which verifies the feasibility and the practicality of the methods above.