Mining spatiotemporal association patterns from complex geographic phenomena

ABSTRACT

Spatiotemporal association pattern mining can discover interesting interdependent relationships among various types of geospatial data. However, existing mining methods for spatiotemporal association patterns usually model geographic phenomena as simple spatiotemporal point events. Therefore, they cannot be applied to complex geographic phenomena, which continuously change their properties, shapes or locations, such as storms and air pollution. The most salient feature of such complex geographic phenomena is the geographic dynamic. To fully reveal dynamic characteristics of complex geographic phenomena and discover their associated factors, this research proposes a novel complex event-based spatiotemporal association pattern mining framework. First, a complex geographic event was hierarchically modeled and represented by a new data structure named directed spatiotemporal routes. Then, sequence mining technique was applied to discover the spatiotemporal spread pattern of the complex geographic events. An adaptive spatiotemporal episode pattern mining algorithm was proposed to discover the candidate driving factors for the occurrence of complex geographic events. Finally, the proposed approach was evaluated by analyzing the air pollution in the region of Beijing-Tianjin-Hebei. The experimental results showed that the proposed approach can well address the geographic dynamic of complex geographic phenomena, such as the spatial spreading pattern and spatiotemporal interaction with candidate driving factors.     

基于空间点模式分析的城市管理事件空间分布及演化——以武汉市江汉区为例

城市网格化管理系统经过多年运行积累了大量历史事件数据, 这类事件数据在空间上呈现明显集聚分布。确定事件发生的空间分布以及衡量空间分布的集聚程度, 能够为城市管理资源的合理调配、划分提供重要的决策支持。本文应用空间点模式分析方法, 对2011 年1-8 月间武汉市江汉区城市网格化管理系统中的两类主体事件(占道经营和垃圾处理类)进行分析, 研究发现:占道经营事件的“热点”区域1-8 月总体呈减少趋势, 而垃圾处理事件的“热点”区域整体呈递增趋势;两类事件呈现明显的空间集聚, 其特征空间尺度都为1000 m左右。研究表明, 空间点模式分析方法能够为城市管理者提供一种针对城市事件空间集聚模式的直观的可视化分析手段, 以及对空间集聚程度的定量分析方法, 并可为进一步统计建模分析奠定基础。     

多源地理大数据支撑下的地理事件建模与分析

地理事件作为描述地理过程的基本单元,逐渐成为地理信息科学（GIS）核心研究内容。由于受人类活动数据获取限制,GIS对地理事件的建模和分析主要关注事件所引起的地理空间要素变化及要素之间的相互影响与作用机制。然而,近年来随着基于位置服务数据（LBS）爆炸式的增长和人类活动大数据定量刻画手段的快速发展,地理事件对人类活动的影响以及公众对地理事件的网络参与度都引起了多个领域的广泛关注,对地理事件的时空认知、建模方法和分析框架提出了巨大的挑战。对此,本文首先深入分析了大数据时代地理事件的概念与分类体系;其次,基于地理事件的时空语义给出了基于图模型的事件数据建模,建立了事件本体及其次生或级联事件的“节点—边”表达结构,开展了事件自身时空演化及其前“因”后“果”的形式化描述;第三,从时空数据分析与挖掘的角度,给出了大数据时代地理事件建模与分析的整体框架,拟突破传统“地理实体空间”事件探测与分析方法的局限性,融合“虚拟空间”事件发现与传播模拟思路,实现多源地理大数据支撑下的面向地理事件的人类活动多尺度时空响应与区域差异分析;最后,本文以城市暴雨事件为例诠释了本文所提出的地理事件建模与分析方法,从城市和城市内部两个尺度进行了暴雨事件与人类活动的一致性响应及区域差异分析,得到了明确的结论,验证了前文分析框架的可行性与实用性。     

Mapping spatiotemporal patterns of events using social media: a case study of influenza trends

Tracking spatial and temporal trends of events (e.g. disease outbreaks and natural disasters) is important for situation awareness and timely response. Social media, with increasing popularity, provide an effective way to collect event-related data from massive populations and thus a significant opportunity to dynamically monitor events as they emerge and evolve. While existing research has demonstrated the value of social media as sensors in event detection, estimating potential time spans and influenced areas of an event from social media remains challenging. Challenges include the unstable volumes of available data, the spatial heterogeneity of event activities and social media data, and the data sparsity. This paper describes a systematic approach to detecting potential spatiotemporal patterns of events by resolving these challenges through several interrelated strategies: using kernel density estimation for smoothed social media intensity surfaces; utilizing event-unrelated social media posts to help map relative event prevalence; and normalizing event indicators based on historical fluctuation. This approach generates event indicator maps and significance maps explaining spatiotemporal variations of event prevalence to identify space-time regions with potentially abnormal event activities. The approach has been applied to detect influenza activity patterns in the conterminous US using Twitter data. A set of experiments demonstrated that our approach produces high-resolution influenza activity maps that could be explained by available ground truth data.     

Exploring spatiotemporal clusters based on extended kernel estimation methods

We examined three different ways to integrate spatial and temporal data in kernel density estimation methods (KDE) to identify space–time clusters of geographic events. Spatial data and time data are typically measured in different units along respective dimensions. Therefore, spatial KDE methods require special extensions when incorporating temporal data to detect spatiotemporal clusters of geographical event. In addition to a real-world data set, we applied the proposed methods to simulated data that were generated through random and normal processes to compare results of different kernel functions. The comparison is based on hit rates and values of a compactness index with considerations of both spatial and temporal attributes of the data. The results show that the spatiotemporal KDE (STKDE) can reach higher hit rates while keeping identified hotspots compact. The implementation of these STKDE methods is tested using the 2012 crime event data in Akron, Ohio, as an example. The results show that STKDE methods reveal new perspectives from the data that go beyond what can be extracted by using the conventional spatial KDE.     

Characterizing and Predicting Traffic Accidents in Extreme Weather Environments

Motorists are vulnerable to extreme weather events, which are likely to be exacerbated by climate change throughout the world. Traffic accidents are conceptualized in this article as the result of a systemic failure that includes human, vehicular, and environmental factors. The snowstorm and concurrent accidents that occurred in the Northeastern United States on 26 January 2011 are used as a case study. Traffic accident data for Fairfax County, Virginia, are supplemented with Doppler radar and additional weather data to characterize the spatiotemporal patterns of the accidents resulting from this major snowstorm event. A kernel density smoothing method is implemented to identify and predict patterns of accident locations within this urban area over time. The predictive capability of this model increases over time with increasing accidents. Models such as these can be used by emergency responders to identify, plan for, and mitigate areas that are more susceptible to increased risk resulting from extreme weather events.     

Analysis of the appearance and disappearance of point objects over time

The appearance and disappearance of immovable points are important spatiotemporal events in geographical information science. They represent phenomena such as the birth and death of trees in forests, construction and destruction of buildings in cities and openings and closures of shops and restaurants. This paper proposes a new method for analyzing the appearance and disappearance of points. The method helps analysts capture the overall picture and regional variation of event pattern and detecting significant local patterns. Four measures are defined that indicate the intensity of spatial and temporal patterns of events. The measures are visualized as grid maps. A statistical test is used to evaluate the significance of the measures to extract the regions of significant patterns. The proposed method is applied in an analysis of shops and restaurants in Shibuya, Tokyo. Technical soundness of the method is discussed along with empirical findings.     

Statistical discrimination of foreshocks from other earthquake clusters

When earthquake activity begins, it may be a foreshock sequence to a larger earthquake, a swarm, or a simple main-shock-aftershock sequence. This paper is concerned with the conditional probability that it will be foreshock activity of a later larger earthquake, depending on the occurrence pattern of some early events in the sequence. The earthquake catalogue of the Japan Meteorological Agency (1926-1993, M_J≥4) is decomposed into a large number of clusters in time and space in order to compare statistical features of foreshocks with those of swarms and aftershocks. Using such a data set, Ogata, Utsu & Katsura (1995) revealed some discriminating features of foreshocks relative to the other types of clusters, for example the events' closer proximity in time and space, and a tendency towards chronologically increasing magnitudes, which encouraged us to construct models which forecast the probability of the earthquakes being foreshocks. Specifically, the probability is a function of the history of magnitude differences, spans between origin times and distances between epicentres within a cluster. For purposes of illustration, the models were fitted to the early part of the data (1926-1975) and the validity of the forecasting procedure was checked on data from the later period (1976-1993). Two procedures for evaluating the performance of the probability forecast are suggested. Furthermore, for the case where only a single event is available (i.e. either it is the first event in a cluster or an isolated event), we also forecast the probability of the event being a foreshock as a function of its geographic location. Then, the validity of the forecast is demonstrated in a similar manner. Finally, making use of the multi-element prediction formula, we show that the forecasting performance is enhanced by the joint use of the information in the location of the first event, and that in the subsequent interevent history in the cluster.     

犯罪热点时空分布研究方法综述

犯罪在地理时空内并不是均匀分布的,而是表现出明显的时空聚集特性,这种聚集性常用“犯罪热点”表述.基于对犯罪热点的理解,从犯罪热点时空分布模式、犯罪热点成因分析以及犯罪热点时空转移及预测等3 个方面总结了当前国内外犯罪热点时空分布相关研究方法的进展.最后,对该领域研究进行了总结与展望.总体上,国内相关研究较少,尚需进一步结合中国国情,提出适用方法.另外,也需要通过相关犯罪理论的深入研究以及其他领域研究方法的借鉴,实现犯罪热点时空分布研究方法的突破与创新.     

基于个体时空邻近指数的户籍维度社会空间分异研究——以上海市郊区10个典型镇为例

社会空间分异是城市研究的经典议题。在人类移动性不断增强的背景下,传统的基于居住空间汇总的社会空间分异测度方法表现出一定的局限性,对于居住空间外的日常活动空间隔离的探讨相对缺乏,对于不同活动和不同时段间分异格局差异的考虑有所不足。因此有必要从“基于人”的视角出发,探索社会空间分异测度的新方法,探讨不同时空间维度的社会分异格局。本研究基于上海市郊区10个典型镇的活动日志调查数据,构建“个体时空邻近指数”,聚焦户籍这一反映中国城市特征的重要维度,以不同户籍类型人群之间的分异程度为研究对象,分析其时空间特征,并对结果进行可视化。研究表明：上海市不同户籍人群在活动、时间和空间维度上存在明显的社会空间分异。本研究创新了基于活动空间的社会空间分异测度方法,从活动与时空间结合的视角探讨了户籍维度的社会空间分异,为更好的理解在中国大城市日益凸显的社会空间分异问题提供了新的视角。     

Take the edge off: A hybrid geographic food access measure

Measures of geographic food access overlook an important source of statistical biases, termed the edge effect. The edge effect refers to the fallacy that events contributing to the spatial pattern of an analysis unit may be outside of that unit; thus merely summarizing events within the unit may lead to distortion of the estimation. Food procurement activities can happen beyond existing administrative boundaries. Delineating food access using unit-based metrics may misrepresent the true space within which food stores are accessible. To overcome this problem, this paper proposes a gravity-based accessibility measure to improve unit-based statistical approaches in food access research. In addition, this method accounts for the spatial interaction between food supply (e.g., food items in stock) and demand (e.g., population) as well as how this interaction is mediated by the spatiotemporal separation (e.g., travel time, modality). The method is applied to the case of Franklin County, OH and has revealed the food access inequity for African Americans by modes of transport, including walking, biking, and driving. The analysis of the correlation between mode-specific food access and socioeconomic status (SES) variables reveals that using a single modality in food access research may not fully capture the travel behavior and its relationship with local food environments. With modifications, the proposed method can help evaluate food access for a target population group, such as Supplemental Nutrition Assistance Program (SNAP) users or selected ethnic minorities who may face acute difficulties in procuring economically affordable and culturally appropriate foods.     

COVID-19期间国家关系交互网络时空分析研究

COVID-19疫情不断蔓延为国际政治、外交关系等带来深刻影响。目前基于复杂网络方法的国际关系研究较少考虑节点的空间属性,难以探索国际关系的动态演化模式及其空间分布特征。该文提出一种结合时间序列聚类与空间统计的国家关系交互网络演化模式探测方法。基于2020年1月-2021年3月的GDELT数据构建国家关系交互网络,基于节点的演化特征,应用K-means聚类算法将节点划分为6种类型,结合局部连接统计方法分析节点演化模式的空间分布特征。研究表明:面对疫情冲击,各国为控制疫情蔓延倾向于参与合作交互事件;国家关系交互网络中的不同时序演化模式总体按照节点的点度中心性强度由高到低分布;疫情防控期间网络中始终处于边缘地位的节点在空间分布上呈现聚集特征,而核心节点空间分布较分散。通过研究网络节点的时序演化模式及空间分布特征可为公共卫生危机事件期间国际关系与地缘政治研究提供新思路,对于危机事件期间制定外交政策与应对策略具有一定参考价值。     

Minimum geocoding match rates: an international study of the impact of data and areal unit sizes

ABSTRACT

The analysis of geographically referenced data, specifically point data, is predicated on the accurate geocoding of those data. Geocoding refers to the process in which geographically referenced data (addresses, for example) are placed on a map. This process may lead to issues with positional accuracy or the inability to geocode an address. In this paper, we conduct an international investigation into the impact of the (in)ability to geocode an address on the resulting spatial pattern. We use a variety of point data sets of crime events (varying numbers of events and types of crime), a variety of areal units of analysis (varying the number and size of areal units), from a variety of countries (varying underlying administrative systems), and a locally-based spatial point pattern test to find the levels of geocoding match rates to maintain the spatial patterns of the original data when addresses are missing at random. We find that the level of geocoding success depends on the number of points and the number of areal units under analysis, but generally show that the necessary levels of geocoding success are lower than found in previous research. This finding is consistent across different national contexts.     

基于气象旱涝指数的旱涝急转事件识别方法

基于长江流域212个气象站点1961-2017年的日降水资料,借助标准化加权平均降水指数（SWAP）,结合多门槛游程理论,提出一种识别旱涝急转事件的新方法。方法应用于旱涝急转事件高发的长江流域,分别从典型站点旱涝事件分析、区域典型旱涝急转事件分析、旱涝急转事件时空分布规律分析等角度,探讨了长江流域1961-2017年旱涝急转事件规律。结论显示：①SWAP指数对于旱涝事件具有良好的识别能力。②聚类方法可聚合相似旱涝急转事件,2011年长江中下游旱涝急转事件中干旱事件占主导地位,持续时间远长于洪涝事件。③ 长江流域旱涝急转事件呈现明显的区域规律：上游发生频率较低,中下游偏高;此外,长江流域多数分区近期旱涝急转事件发生频率呈现上升趋势。研究结果表明,基于SWAP指数并结合多门槛游程理论的方法能够比较准确地识别旱涝急转事件,可进一步应用于旱涝急转事件的预测及评估中。     

Attacks on oil transport pipelines in Nigeria: A quantitative exploration and possible explanation of observed patterns

In countries with insufficient investments in infrastructure and weak environmental governance, oil leakage from pipelines often occurs as a result of poor management and maintenance. Nigeria has its share of such incidents, but also, it suffers a large number of deliberate attacks (‘interdictions’) on oil pipelines. Often these attacks are accompanied by oil theft, carried out by well-equipped professionals and/or at a smaller scale by opportunistic local residents. The causes of these attacks, and the extent of subsequent damage to local communities and the environment, are obscured by a complex web of stakeholders, claims and actions. Any efforts to mitigate the negative impacts of interdiction on the environment and people require a better understanding of its spatiotemporal pattern of occurrence. This article presents a first quantitative and regional exploration of the problem of oil pipeline interdiction in Nigeria. It illustrates geographic patterns through choroplethic and bivariate GIS (Geographical Information Systems) map overlays. We examine interdiction statistics, identify spatiotemporal patterns and discuss correlations with socioeconomic factors. Findings include: (a) strong negative correlation between pipeline interdiction and poverty; and (b) statistically and non-statistically significant mean differences in the pattern of interdiction occurrence amongst the five geographic regions. Finally, we highlight the need for much better data collection and reporting for the mitigation of the negative socio-environmental impacts of interdiction incidences.     

An approach for understanding human activity patterns with the motivations behind

A mechanistic understanding of human activity patterns lays a foundation for many applications. The majority of the current research aims to outline human activity patterns mainly from spatiotemporal perspectives (i.e., modeling human mobility patterns), lacking of understanding of the motivations behind behaviors. The aim of this study is to model and understand human activity patterns within urban areas using both spatiotemporal and cognitive psychology methods to measure both human behavior patterns and the underlying motivations . We first propose a framework that enables us to analyze the spatiotemporal patterns of urban human activities, infer the associated semantic patterns that represent the motivations driving human mobility choices and behaviors, and measure the similarity between human activities. We then construct a human activity network based on the similarity to depict human activity patterns. The framework is applied to a case study of Toronto, Canada, where geotagged tweets are used as a proxy for human activities to explore activity patterns. The analysis of the human activity network shows that 61% of tweeter users follow similar activity patterns. Our work provides a new tool for better understanding the way individuals interact with urban environments that could be applied\ to a variety of urban applications.     

Predicting potential distributions of geographic events using one-class data: concepts and methods

One common problem with geographic data is that, for a specific geographic event, only occurrence information is available; information about the absence of the event is not available. We refer to these specific types of geospatial data as geographic one-class data (GOCD). Predicting the potential spatial distributions that a particular geographic event may occur from GOCD is difficult because traditional binary classification methods that require availability of both positive and negative training samples cannot be used. The objective of this research is to define GOCD and propose novel approaches for modelling potential spatial distributions of geographic events using GOCD. We investigate the effectiveness of one-class support vector machine (OCSVM), maximum entropy (MAXENT) and the newly proposed positive and unlabelled learning (PUL) algorithm for solving GOCD problems using a case study: species distribution modelling from synthetic data. Our experimental results indicate that generally OCSVM, MAXENT and PUL are effective in modelling the GOCD. Each method has advantages and disadvantages, but PUL seems to be the most promising method.     

基于局部运动向量时空统计的极光事件检测

极光动态过程的分析与理解对极光发生机制研究具有重要意义。本文提出了一种基于动态过程的极光事件检测方法。首先利用多尺度流体光流的方法提取出极光的局部运动场信息,然后基于局部运动场时空统计特性表征极光视频序列,最后实现对特殊极光事件的检测。实验结果表明,本文方法能够高效、准确地检索到特殊极光事件,并且检测结果不依赖于目标事件的选择。这一成果为开展基于大量连续观测的极光视频对极光动态过程进行统计分析的研究奠定了基础。     

Aoristic analysis: the spatial interpretation of unspecific temporal events

Temporal limitations of GIS databases are never more apparent than when the time of a change to any spatial object is unknown. This paper examines an unusual type of spatiotemporal imprecision where an event occurs at a known location but at an unknown time. Aoristic analysis can provide a temporal weight and give an indication of the probability that the event occurred within a defined period. Visualisation of temporal weights can be enhanced by modifications to existing surface generation algorithms and a temporal intensity surface can be created. An example from burglaries in Central Nottingham (UK) shows that aoristic analysis can smooth irregularities arising from poor database interrogation, and provide an alternative conceptualisation of space and time that is both comprehensible and meaningful.