摘要
Web文本中蕴含着丰富的以自然语言描述的非结构灾害信息和知识。基于Web文本自动提取和构造结构化、综合性灾害信息,是灾害信息领域研究的前沿问题。目前国内外利用Web文本挖掘技术在灾时与灾后的应急响应与救援,灾害的早期预警和风险分析方面进行示范应用;同时在文本灾害信息的语义理解与抽取、文本灾害信息的时空匹配、以及文本灾害信息的不确定性和可靠性评价等关键技术领域迅速展开研究。我国应加强以Web文本为信息源的中文灾害信息挖掘关键技术、软件,以及管理体系的研究,以有效弥补灾害研究与管理过程中灾害数据共享困难,以及可利用的动态实时、综合性灾害数据缺乏的薄弱环节,提升灾害信息服务水平。
Web is full of disaster information and knowledge,which is non-structural and described in natural language.Automatic extraction and construction of structural,comprehensive disaster information from Web is an advanced issue in the fields of disaster information.Currently,some information systems of disaster data extraction from Web have been applied in disaster emergency response and rescue,disaster early warning and risk analysis.Meanwhile,several key techniques such as semantic understanding of Web pages and information extraction,temporal-spatial matching of disaster information,and uncertainty and reliability evaluation of disaster information,are emphasized.The research of key technology,software and administrative management of disaster information extraction from web should be enhanced in China,which would effectively overcome the difficulties of information sharing in disaster research and management and cover the shortage of dynamic,real time and comprehensive information of disasters.
出处
《灾害学》
CSCD
2010年第2期119-123,128,共6页
Journal of Catastrophology
基金
上海市教委重点学科<地理学与城市环境>(J50402)
上海师范大学科研项目(SK200846)
上海师范大学重点培育学科项目(DZL801)
上海市科委科技攻关项目(08240514000)
关键词
WEB文本
灾害信息
空间信息
挖掘技术
Web pages
disaster information
spatial information
data extraction
