In this paper,a method is proposed for query for location of resource science and technology information in large data.Firstly,chaos theory was used for reference to rebuild phase space of attribute character of resource information and invariant of original geometrical characteristic of resource data was worked out.Then delay time of phase-space reconstruction was provided and optimal embedded dimension was searched.Moreover,corresponding feature vector of the information was extracted and clustered,and multi-population glowworm theory was introduced into process of the fast query of information location to provide fitness function.The simulation show that the method has high precision of information extraction and can lay foundation for improving quality of large data.%对大数据中资源科技信息进行快速查询,能够提升大数据的质量.对资源信息的定位快速查询,需要提取资源科技信息的关联特征向量,给出资源科技信息定位快速查询适应度函数,完成对资源科技信息定位的快速查询.传统方法组建数据特征的不同尺度空间矩阵,提取资源科技信息特征,但忽略了求取适应度函数,导致资源信息查询精度偏低.所提方法借鉴了混沌理论思想重构资源信息属性特征的相空间,计算出资源数据原始几何特征不变量,给出相空间重构的延迟时间,搜素最佳嵌入维数,提取资源科技信息的关联特征向量,并进行聚类,将多子群萤火虫理论引入到对大数据中资源科技信息定位快速查询过程中,给出大数据中资源科技信息定位快速查询适应度函数,完成对大数据中资源科技信息定位快速查询.仿真证明,所提方法信息抽取精度高,可以为提升大数据的质量奠定了基础.
展开▼