互联网“泛在”空间数据源中全球“聚焦”式点要素萃取和可视化关键技术

With the need of globalization of military, diplomacy, economy, environment protection and many other areas in our nation, the demand for global geographic information, especially for overseas geographic information, is becoming more and more urgent. In recent years, the maturity of online map services, combination and popularity of LBS and social networks, the sharing of crowdsourcing data and volunteered geographic information (VGI) have made it possible to obtain large-scale global vector data from a wide range of data sources on the Internet that directly or indirectly are related to geographic information (simply as "ubiquitous" spatial data sources)..Point geographic feature (such as place names, points of interest, etc.) is one of the most important and basic types within the global geographic information framework. But the traditional description method of “Geometry + Attributes” can hardly meet people’s demands for a full range of perception and expression of the point features in the real world. On the other hand, there is a large amount of information such as the geometric position, the name, the encyclopedia document, the picture, the location, the microblog, the video and other information of the point features in the “ubiquitous” spatial data sources, but structured and unstructured, spatial and non-spatial data coexists, information and data garbage are mixed, their qualities varies greatly, only after "the sweeping of great waves” can it form a global point feature set that “Focus” tens of millions pieces of information(simply as global "focus" point geographic features). Then through visualization, it takes a map form that fits the spatial cognition result expression. Users can not only get a specific focus on multiple perspectives and real-time depth theme knowledge from a micro cosmic point of view, but also obtain cognition of the situation of large regions or the globe, and even the future trend from a macro cosmic point of view..We propose a new method which can extract and visualize global "focus" geographic point features from "ubiquitous" spatial data sources on the Internet, including three aspects:.(1).Extraction mechanisms of worldwide "ubiquitous" spatial data sources;.(2).How to design Global focused model about point features and establish a classification based on the impact factor index;.(3).More than millions number of "focus" type point features visualization and usability evaluation.Researching on this topic will help to improve our ability to acquire global geospatial information from the Internet greatly and provide critical technical means of support for the construction of a global multi-scale vector database. It can avoid the dilemma of "no big data available" in the era of big data, help to compensate for the lack of space cognitive mapping theory on the global "focus" type point features of visual direction and crack the "Exaflood" problem, digging "small" entity wisdom from the "big" data.

随着国家军事、外交、经济、环保等诸多领域全球化发展的需要，我国对全球地理信息，尤其是境外地理信息的需求日益迫切。本项目提出一种从互联网中广泛存在、分布无序、直接或间接与地理信息相关的数据源中（简称为“泛在”空间数据源）获取全球范围内的点要素及其相关信息的一种新思路，并构建以点为中心的多重视角和实时纵深的信息模型（简称为“聚焦”式点要素模型），在此基础上设计新的适宜空间认知的可视化方法，从而满足人们对现实世界中关于某个点状要素的全方位感知与表达的需求。.开展本项目的研究将有助于大幅提高我国从互联网中获取全球地理空间信息的能力，为我国全球多尺度矢量数据库的构建提供关键技术手段的支撑，避免大数据时代“无大数据可用”的窘境，同时有助于弥补全球“聚焦”式点要素可视化方向上空间认知制图理论的不足，破解大数据时代“数字洪水”的难题，从“大”数据中挖掘“小”实体的智慧。

随着国家军事、外交、经济、环保等诸多领域全球化发展的需要，我国对全球地理信息，尤其是境外地理信息的需求日益迫切，而从互联网中获取全球地理信息是一种重要的途径。.本项目以互联网中专业地理数据部门、在线地图服务供应商以及众包地理数据源等“泛在”空间数据源为研究对象，针对不同的数据源研制了相应的爬虫软件，重点突破了位置照片、商用地图网站POI点的全球范围内大规模获取技术以及位置文本的识别和抽取技术，获得了全球范围内TB级地理空间数据，数据项达亿级，建立了“聚焦”式点要素模型和相应的数据库，并在校园网内搭建互联网全球地理数据检索与共享平台。最后在整合和处理后的互联网位置数据基础上，将Cartogram表示方法进一步扩展，提出了基于移动最小二程法的时间距离可视化表达方法，个人地理标记数据的可视化方法和基于标签云的位置关联文本信息可视化方法并进行了可用性评价，为突破百万级数量以上的点要素可视化效率瓶颈而对热力图生成算法做了优化和改进，并以旅游景点的评价和城市活力度度量为应用做了进一步的数据分析和挖掘。.结合本项目的研究进展，项目组发表期刊论文12篇，已录用期刊论文1篇，其中EI源期刊5篇，1篇发表在德国制图学杂志《Kartographische Nachrichten》（英文名：Journal of Cartography and Geographic Information，Scopus检索），北大中文核心期刊5篇；获得国家发明专利1项，1项国家发明专利在实质审查阶段；参加1次国际学术会议（瑞士苏黎世）；完成硕士学位论文1篇，1篇博士学位论文正在撰写中（2018年6月毕业）；形成了2份内部报告；研发的互联网全球地理数据检索与共享平台已在校园网内部提供空间数据共享和整合服务。这些成果的取得，为我国全球多尺度矢量数据库的构建，境外制图以及一带一路的相关建设提供了重要的数据参考和关键技术支撑。