欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
应急知识模式是关于突发事件演进规律的知识，因而是应急领域知识库的重要组成部分。应急知识模式提取有助于应急工作者有效了解事态现状和快速做出应急反应，进而降低事件负面影响。本文研究如何从海量应急领域文本中抽取应急工作者关注的知识模式。我们用自然语言处理技术（NLP）从无结构海量应急领域文本中提取描述突发事件的特征事实，并通过提取特征事实之间的因果推理关系和并列作用关系，来构建知识的特征事实依赖图（FEDG）模型。基于FEDG模型，我们提出使用闭合约束搜索算法提取应急知识模式。通过实验及其分析证明本方法对知识模式提取具有较好效果。
英文摘要：
Emergency knowledge pattern expresses the development direction of emergencies, thus is one of the most important parts in the domain knowledge base. The extraction of emergency knowledge pattern is helpful when decision makers need to get the view of current situations and make rapid response. This paper presents a method to extract knowledge patterns from mass texts in emergency field. Feature Events are extracted using Natural Language Process （NLP） technical from unstructured texts and the dependency relation between Feature Events are described using FEDG model. The knowledge patterns are extracted based on this model using Double Closed Constrained Search Algorithm. The result of our experiment proved that the method of extracting knowledge pattern is effective and reliable.
同期刊论文项目
同项目期刊论文
期刊信息
位置：