基于语义类型本体的地名组配查询

地名查询方式多种多样,但它们都没有考虑地名语义类型因素,而语义类型恰恰是地名信息中相当重要的一部分。以地名语义分类为基础,构建了地名语义类型本体模型,并以郑州市为例进行了地名组配查询实验。结果表明,该方式对于检索地名的相关信息具有较好的辅助作用,为地名查询与检索作了有益的补充。     

需求驱动兵团空间数据本体语义的描述

针对兵团空间信息集成与共享异构数据语义的广泛性和复杂性,本文依据其领域特性和共享需求,采用本体作为兵团空间数据表示的方法,从应用领域出发,以需求来驱动兵团空间数据本体框架构建,提出了兵团空间数据的本体工程构建方法,研究了本体领域、概念类型、语义关系,并对空间数据本体进行了语义描述。评估表明该方法构建的兵团空间数据本体有效,有利于数据共享与互操作。     

Mapping collective human activity in an urban environment based on mobile phone data

Identifying and characterizing variations of human activity – specifically changes in intensity and similarity – in urban environments provide insights into the social component of those eminently complex systems. Using large volumes of user-generated mobile phone data, we derive mobile communication profiles that we use as a proxy for the collective human activity. In this article, geocomputational methods and geovisual analytics such as self-organizing maps (SOM) are used to explore the variations of these profiles, and its implications for collective human activity. We evaluate the merits of SOM as a cross-dimensional clustering technique and derived temporal trajectories of variations within the mobile communication profiles. The trajectories’ characteristics such as length are discussed, suggesting spatial variations in intensity and similarity in collective human activity. Trajectories are linked back to the geographic space to map the spatial and temporal variation of trajectory characteristics. Different trajectory lengths suggest that mobile phone activity is correlated with the spatial configuration of the city, and so at different times of the day. Our approach contributes to the understanding of the space-time social dynamics within urban environments.     

一种新的近似平行沟渠群的典型化方法

提出并实现了一种针对近似平行沟渠群的典型化方法。首先,建立沟渠的方位关系图,通过生成关系图中各连通分量的极大完全子图,提取近似平行的沟渠;然后,依据视觉认知规律定义沟渠间的认知距离,并基于最小生成树的剪枝对沟渠进行分组;接着,运用K-means ++算法对分组后的沟渠进行重新表达;最后,通过比较综合前后沟渠群的外轮廓相似性以及沟渠分布密度的差异,对典型化结果进行了评价。对广州市数据进行了实验,结果表明该方法能有效进行近似平行的沟渠群的典型化。     

从大模型看测绘时空信息智能处理的机遇和挑战

当前,时空信息、定位导航服务已成为重要的新型基础设施,在通用人工智能的驱动下,大模型引领的智能时代已经到来,越来越强大的大模型将在测绘时空信息智能处理与应用中发挥越来越重要的作用。以大模型为研究范式,总结大模型在测绘时空信息智能处理的现状与进展,分析大模型在测绘时空信息智能处理面临的挑战,阐述多模态融合与理解架构设计、提示工程优化微调以及人在回路引导决策3个核心方面在时空信息测绘大模型中的关键作用。针对行业理解深度、数据安全隐患、内容可信度保障以及训练部署成本优化4个方面,展望时空信息测绘大模型面临的挑战与发展趋势。

     

基于FME语义映射的数据转换及应用

在详细介绍FME语义映射和转换规则的基础上,分析了FME的空间数据转换机制和简单语义转换规则的不足,并基于自定义映射文件提出了一种在DOS环境下进行批量数据转换的高效方法,最后以DGN到DWG格式的数据转换实例探讨了实施FME语义转换的具体步骤。     

基于语义相似度的海量遥感图像可视化研究

针对海量遥感图像如何有效的传达分类结果以实现有效的可视化问题,本文按照分类的语义标注结果的相似度并运用现有的信息可视化技术来实现图像的可视化。首先采用了贝叶斯网络学习的方法进行图像的自动分类标注,然后利用基于图像布局的多维标度算法(Multi-dimensional Scale)以及无需降维的Value and Relation(VaR)技术实现可视化。实验表明本文的方法能够填补图像低层视觉特征和高层语义之间鸿沟,对大量的图像在一个视图内进行有效的浏览,而不造成图像的混乱,并能实现高层次的图像分析。实验的可视化结果是十分有效的。     

基于模糊理论的时空数据模型及算法解决

针对当前时空数据模型中存在的主要问题,本文提出了一种基于模糊理论的时空数据模型.该模型从时空对象演化的角度出发,结合模糊隶属度的概念,力求精确表达对象演化的整个过程及对象之间的相互关系.在此基础上,进一步分析了该模型的数据结构以及算法解决,为时空数据模型的研究提供了新的思路.     

利用语义的多版本网络覆盖服务模式匹配方法

提出了一种基于节点相似度的语义模式匹配方法,首先对元素标签进行概念化,其次计算元素标签间的语义关系,再次计算节点概念并转化成合取范式CNF,最后利用命题可满足性问题解决器SAT验证节点间的概念关系。实验结果表明,该方法有效提高了匹配的查全率、精度和整体质量。     

Geostatistical approaches to refinement of digital elevation data

Data refinement refers to the processes by which a dataset’s resolution, in particular, the spatial one, is refined, and is thus synonymous to spatial downscaling. Spatial resolution indicates measurement scale and can be seen as an index for regular data support. As a type of change of scale, data refinement is useful for many scenarios where spatial scales of existing data, desired analyses, or specific applications need to be made commensurate and refined. As spatial data are related to certain data support, they can be conceived of as support-specific realizations of random fields, suggesting that multivariate geostatistics should be explored for refining datasets from their coarser-resolution versions to the finer-resolution ones. In this paper, geostatistical methods for downscaling are described, and were implemented using GTOPO30 data and sampled Shuttle Radar Topography Mission data at a site in northwest China, with the latter’s majority grid cells used as surrogate reference data. It was found that proper structural modeling is important for achieving increased accuracy in data refinement; here, structural modeling can be done through proper decomposition of elevation fields into trends and residuals and thereafter. It was confirmed that effects of semantic differences on data refinement can be reduced through properly estimating and incorporating biases in local means.