面向遥感影像需求自然语言描述解析与推理的时间本体建模

Time Ontology Modeling for Remote Sensing Image Requirements of Natural Language Descriptive Analysis and Ratiocination

随着遥感影像数据的激增,如何从海量遥感数据中检索出用户所需要的目标数据,是遥感影像服务所面临的关键问题之一。现有的遥感影像搜索服务大多是基于关键字的遥感影像元数据检索,还不支持遥感应用领域的自然语言需求理解搜索,在信息的查全率、查准率等方面都无法满足用户日益增长的需求。面向遥感影像需求自然语言描述的解析及推理,对遥感影像需求的时间信息在表达、识别、抽取、建模方面进行探讨,构建了遥感影像需求的时间本体。原型系统的实验表明该本体模型适用于遥感影像需求的时间解析与推理,能提升遥感影像智能服务的效果。

Along with the soar of remote sensing image data,how to retrieve out user needed target data from massive remote sensing data is one of key issues the remote sensing service faced.Available remote sensing image search services are mostly based on keyword retrieval of remote sensing image elemental data,and nonsupport to remote sensing application field natural language requirement understanding search,in aspects of information recall ratio and precision ratio all unable to meet users growing demand.Faced remote sensing image requirements of natural language descriptive analysis and ratiocination have discussed remote sensing image required time information from expression,identification,extraction and modeling aspects,established remote sensing requirement time ontology.Prototype system experiment has shown that this ontology model is adaptable to remote sensing image required time analysis and ratiocination,can promote remote sensing image intellectualized service effect.