面向服务的地学多源数据虚拟整合及其可视化分析

科学数据共享的推进,给科研人员带来了前所未有的科研契机,但是依然没有摆脱"搜集数据-下载数据-整理分析数据"的科研流程,这种传统的科研流程已经严重阻碍了科研效率的产出,科研人员对信息和知识的需求已经远远大于对数据本身的需求,而要实现对信息和知识的挖掘,多源数据的实时整合与在线可视化分析是其关键。本文选择社会经济研究主题,以中国社会经济统计数据、国家基础地理行政区划矢量数据为例,并结合ESRI和Google的全球地图与影像服务,采用在线虚拟整合的方法,探讨了面向服务的地学多源数据集成方案及其可视化分析,以期达到快速的知识发现与信息获取。研究结果表明:面向服务的技术架构虽然已经成熟,相对于传统的数据共享和应用系统在数据接口开放性、集成性和应用扩展性等方面具有明显的优势,能够切实解决多源异构数据的整合问题,但是技术的成熟并不等于应用的成熟,广泛而成熟的应用不但需要技术的支撑,更需要友好和智能的操作界面,这方面还需要较长的时间去探索。从本文的应用实例来看,在面向服务技术的深入应用过程中,还存在着不同来源数据单位不统一、操作流程复杂、服务规范不统一等问题。同时,在不同类型服务的标准化、操作智能化和多源服务的功能应用模块化等方面还需要大量的实践,才能进一步推进地学信息化科研环境的建设,以全方位为科研人员服务。

A Service-oriented Virtual Integration and Visualization Analysis of Geo-multisource Data

Although the promotion of scientific data sharing has brought an unprecedented research opportunity,it still did not get rid of the traditional research process that is ＂collecting data-downloading data-analyzing data＂,which has severely hampered the efficiency of research output and the needs of data for researchers has been changed into the needs of information and knowledge.In this paper,taking China＇s socio-economic statistical data,vector data of national fundamental geographic administrative divisions as examples,in conjunction with ESRI and Google＇s global maps and image services,we discussed the virtual data integration methods of geo-multisource data and its visualization analysis application,so as to realize rapid knowledge discovery and information acquisition.The paper firstly described methods to publish the map data and attribute data as web services and then discussed three integration principles of multisource web services for the heterogeneous and homogeneous data：（1） all the records of attribute data are assigned a unique number as key fields for the association with a spatial map shape,so that the spatialization of attribute data could be realized;（2） all the spatial data are preprocessed to be suitable for integration by unifying their projections and coordinate systems;and（3） the integrity and accuracy of the data published for web services are ensured.Following the above three principles,the data were prepared.After publishing different types of data as web services,a scheme of online data statistical analysis and visualization based on integration data retrieved from multi-source web services were designed and a preliminary application system was developed.The results show that service-oriented technology can effectively solve the problems of multisource heterogeneous data integration and have obvious advantages as opposed to traditional data sharing and application system.Although the service-oriented technology has turned to be mature,there is still a long way to go for the wide and deep applications.According to our study,there still exist some problems such as the different data units from different sources,the complex operation processes and the ununiform service specification.In future studies,more attention should be paid to the standardization of different services,intelligent operations and application packaging,so as to promote the construction of E-Geoscience and provide all-round services for researchers ultimately.