How where is when? On the regional variability and resolution of geosocial temporal signatures for points of interest

The temporal characteristics of human behavior with respect to points of interest (POI) differ significantly with place type. Intuitively, we are more likely to visit a restaurant during typical lunch and dinner times than at midnight. Aggregating geosocial check-ins of millions of users to the place type level leads to powerful temporal bands and signatures. In previous work these signatures have been used to estimate the place being visited based purely on the check-in time, to label uncategorized places based on their individual signature's similarity to a type signature, and to mine POI categories and their hierarchical structure from the bottom up. However, not all hours of the day and days of the week are equally indicative of the place type, i.e., the information gain between temporal bands that jointly form a place type signature differs. To give a concrete example, places can be more easily categorized into weekend and weekday places than into Monday and Tuesday places. Nonetheless, research on the regional variability of temporal signatures is lacking. Intuitively, one would assume that certain types of places are more prone to regional differences with respect to the temporal check-in behavior than others. This variability will impact the predictive power of the signatures and reduce the number of POI types that can be distinguished. In this work, we address the regional variability hypothesis by trying to prove that all place types are created equal with respect to their temporal signatures, i.e., temporal check-in behavior does not change across space. We reject this hypothesis by comparing the inter-signature similarity of 321 place types in three major cities in the USA (Los Angeles, New York, and Chicago). Next, we identify a common core of least varying place types and compare it against signatures extracted from the city of Shanghai, China for cross-culture comparison. Finally, we discuss the impact of our findings on POI categorization and the reliability of temporal signatures for check-in behavior in general.     

基于地理邻近性的自编码器在地点推荐中的应用

个性化地点推荐系统对于基于位置的社交网络（Locationbased Social Networks, LBSNs）的发展至关重要。它不仅能够帮助用户挖掘新的地点,同时也有利于服务商更好地提供个性化服务。现存关于这方面的研究,将所有的地点同等看待。但是在不同类别中,签到频率的数据规模却不可同等看待。本文基于TFIDF理论将签到频率转换成基于类别的偏好数据,提出一个基于地理邻近性的深度自编码器模型,利用签到数据中的地理信息构造推荐系统。在LBSNs真实数据集上进行实验分析,结果表明相对于对比算法,本文模型的实验结果更好,基于地理邻近性的深度自编码器模型适用于地点推荐任务。     

Assessing the influence of point-of-interest features on the classification of place categories

Points of interest (POIs) digitally represent real-world amenities as point locations. POI categories (e.g. restaurant, hotel, museum etc.) play a prominent role in several location-based applications such as social media, navigation, recommender systems, geographic information retrieval tools, and travel-related services. The majority of user queries in these applications center around POI categories. For instance, people often search for the closest pub or the best value-for-money hotel in an area. To provide valid answers to such queries, accurate and consistent information on POI categories is an essential requirement. Nevertheless, category-based annotations of POIs are often missing. The task of annotating unlabeled POIs in terms of their categories — known as POI classification — is commonly achieved by means of machine learning (ML) models, often referred to as classifiers. Central to this task is the extraction of known features from pre-labeled POIs in order to train the classifiers and, then, use the trained models to categorize unlabeled POIs. However, the set of features used in this process can heavily influence the classification results. Research on defining the influence of different features on the categorization of POIs is currently lacking. This paper contributes a study of feature importance for the classification of unlabeled POIs into categories. We define five feature sets that address operation based, review-based, topic-based, neighborhood-based, and visual attributes of POIs. Contrary to existing studies that predominantly use multi-class classification approaches, and in order to assess and rank the influence of POI features on the categorization task, we propose both a multi-class and a binary classification approach. These, respectively, predict the place category among a specified set of POI categories, or indicate whether a POI belongs to a certain category. Using POI data from Amsterdam and Athens to implement and evaluate our study approach, we show that operation based features, such as opening or visiting hours throughout the day, are the most important place category predictors. Moreover, we demonstrate that the use of feature combinations, as opposed to the use of individual features, improves the classification performance by an average of 15%, in terms of F1-score.     

基于多源地理大数据的城市空间结构研究

以东莞市主城区为研究区,利用夜光遥感数据、POI数据与手机定位强度数据,采用核密度分析、数据格网化与双因素组合制图方法,获得3种数据空间耦合相同或相异区域,并比较其与城市空间结构的关系.研究表明,3种数据的空间分布趋势总体一致,部分区域出现空间耦合相异:①受交通、功能区与夜光遥感数据的"溢出"效应等因素影响,城市道路、...     

Spatial co-location pattern mining for location-based services in road networks

With the evolution of geographic information capture and the emergency of volunteered geographic information, it is getting more important to extract spatial knowledge automatically from large spatial datasets. Spatial co-location patterns represent the subsets of spatial features whose objects are often located in close geographic proximity. Such pattern is one of the most important concepts for geographic context awareness of location-based services (LBS). In the literature, most existing methods of co-location mining are used for events taking place in a homogeneous and isotropic space with distance expressed as Euclidean, while the physical movement in LBS is usually constrained by a road network. As a result, the interestingness value of co-location patterns involving network-constrained events cannot be accurately computed. In this paper, we propose a different method for co-location mining with network configurations of the geographical space considered. First, we define the network model with linear referencing and refine the neighborhood of traditional methods using network distances rather than Euclidean ones. Then, considering that the co-location mining in networks suffers from expensive spatial-join operation, we propose an efficient way to find all neighboring object pairs for generating clique instances. By comparison with the previous approaches based on Euclidean distance, this approach can be applied to accurately calculate the probability of occurrence of a spatial co-location on a network. Our experimental results from real and synthetic data sets show that the proposed approach is efficient and effective in identifying co-location patterns which actually rely on a network.     

DynamiCITY: Revealing city dynamics from citizens social media broadcasts

Social media and mobile devices have revolutionized the way people communicate and share information in various contexts, such as in cities. In today’s “smart” cities, massive amounts of multiple forms of geolocated content is generated daily in social media, out of which knowledge for social interactions and urban dynamics can be derived. This work addresses the problem of detecting urban social activity patterns and interactions, by modeling cities into “dynamic areas”, i.e., coherent geographic areas shaped through social activities. Social media users provide the information on such social activities and interactions in cases when they are on the move around the city neighborhoods. The proposed approach models city places as feature vectors which represent users visiting patterns (social activity), the time of observed visits (temporal activity), and the context of functionality of visited places category. To uncover the dynamics of city areas, a clustering approach is proposed which considers the derived feature vectors to group people’s activities with respect to location, time, and context. The proposed methodology has been implemented on the DynamiCITY platform which demonstrates neighborhood analytics via a Web interface that allows end-users to explore neighborhoods dynamics and gain insights for city cross-neighborhood patterns and inter-relationships.     

空间位置的关联分析及其向量化表示方法

理解地理空间位置的空间相关性,对于地理信息检索、推荐系统,城市交通管理,居民出行模式探究等应用研究具有重要支撑作用.为更具体表义空间位置及其关联关系,本文基于多种居民出行轨迹数据,提出一种基于深度学习的空间位置向量化表示方法,而后通过空间位置向量的向量运算,可计算得到空间位置的关联程度.首先将长、短距离出行轨迹进行匹配连接,构建大规模交通网络,覆盖多种出行模式,得到对不同位置间空间关联信息的完整识别.然后基于图神经网络模型,本文提出融合位置特征与轨迹信息的空间向量化表示方法,并优化其训练学习中节点采样方法,提高空间向量的表达能力.最后以北京市共享单车轨迹数据与公共交通路网数据进行实证分析,实验结果表明基于本文提出方法生成的空间向量在空间位置的关联分析、聚类分析中相比DeepMove等已有方法拥有更好的效果.     

Evaluating a location distortion model to improve reverse geocoding through temporal semantic signatures

Reverse geocoding is a process that maps coordinates to a set of location identifiers such as addresses or toponyms. What makes the reverse geocoding process challenging is the uncertainty of the position being asked and the point features used to represent places. In recent years, due to advances in locating technologies, large amounts of location-based data have been produced in location-based social networks such as the Yelp, Foursquare, and Swarm. These data are a rich source of information about the patterns of people's behaviors in different places. In this paper, with the help of these data, the enhancement of spatial distance-only reverse geocoding has been attempted. The main purpose of this paper is to develop and validate an algorithm for matching categories in the Yelp and Swarm services. In this way, the data from the Yelp were used for generating temporal behavior data and the data from Swarm were used for collecting check-in data. Since the data from Yelp and Swarm services have different categorization structures, integrating these two structures was one of the main challenges of our study. After matching the categories of Yelp and Swarm services, the obtained temporal behavior data for all data sets of Yelp were used in the process of reverse geocoding for Swarm check-in data. In our study, linear, rational and sinusoidal functions were used for distorting the spatial distance with temporal check-in probability in the process of reverse geocoding. In addition, two sets of data include training and test data were used for determining the parameters of the model and validating the results. In this way, it was found that by combining a linear model with temporal behavior data, the results of spatial distance-only reverse geocoding can be improved by 29.96% for the Mean Reciprocal Rank index (a statistical measure for evaluating any process that produces a list of responses, ordered by probability of correctness) and 105.73% for the First Position index (which counts the number of correctly identified POIs). The findings of our study confirmed that the extended set of temporal probabilities of POI categories obtained from Yelp and Swarm gives better results than previous studies. The strengths of our method was demonstrated by validating it against a spatial distance only baseline by the Mean Reciprocal Rank and the First Position indices.     

Beyond the bar: the places where location-based services are used in the city

Mobile services are integrating into the places and routines of daily life. But which types of places afford the use of various services, and how important are these places in our lives? Through several studies, we have explored the types of places that are most important to people in their cities, and compare these to the place types where different location-based services are used. We find that services were used quite consistently between cities, but that between services places of personal salience, such as parks, are less common in the use of today’s check-in services compared with location-based storytelling systems. Supported with data from the StoryPlace.me service, we suggest that focusing on selective sharing and storytelling can facilitate use at these more personally meaningful places.     

Optimal group route query: Finding itinerary for group of users in spatial databases

The increasing popularity of location-based applications creates new opportunities for users to travel together. In this paper, we study a novel spatio-social optimization problem , i.e., Optimal Group Route, for multi-user itinerary planning. With our problem formulation, users can individually specify sources and destinations, preferences on the Point-of-interest (POI) categories, as well as the distance constraints. The goal is to find a itinerary that can be traversed by all the users while maximizing the group’s preference of POI categories in the itinerary. Our work advances existing group trip planning studies by maximizing the group’s social experience. To this end, individual preferences of POI categories are aggregated by considering the agreement and disagreement among group members. Furthermore, planning a multi-user itinerary on large road networks is computationally challenging. We propose two efficient greedy algorithms with bounded approximation ratio, one exact solution which computes the optimal itinerary by exploring a limited number of paths in the road network, and a scaled approximation algorithm to speed up the dynamic programming employed by the exact solution. We conduct extensive empirical evaluations on two real-world road network/POI datasets and our results confirm the effectiveness and efficiency of our solutions.     

Mining spatial colocation patterns: a different framework

Recently, there has been considerable interest in mining spatial colocation patterns from large spatial datasets. Spatial colocation patterns represent the subsets of spatial events whose instances are often located in close geographic proximity. Most studies of spatial colocation mining require the specification of two parameter constraints to find interesting colocation patterns. One is a minimum prevalent threshold of colocations, and the other is a distance threshold to define spatial neighborhood. However, it is difficult for users to decide appropriate threshold values without prior knowledge of their task-specific spatial data. In this paper, we propose a different framework for spatial colocation pattern mining. To remove the first constraint, we propose the problem of finding N-most prevalent colocated event sets, where N is the desired number of colocated event sets with the highest interest measure values per each pattern size. We developed two alternative algorithms for mining the N-most patterns. They reduce candidate events effectively and use a filter-and-refine strategy for efficiently finding colocation instances from a spatial dataset. We prove the algorithms are correct and complete in finding the N-most prevalent colocation patterns. For the second constraint, a distance threshold for spatial neighborhood determination, we present various methods to estimate appropriate distance bounds from user input data. The result can help an user to set a distance for a conceptualization of spatial neighborhood. Our experimental results with real and synthetic datasets show that our algorithmic design is computationally effective in finding the N-most prevalent colocation patterns. The discovered patterns were different depending on the distance threshold, which shows that it is important to select appropriate neighbor distances.     

基于距离类别的多源兴趣点融合算法

为了更好地实现多源兴趣点（POI）数据的有效集成与精确融合,提出了一种结合空间与非空间属性的距离类别的兴趣点融合算法（MNMDC）。首先,对空间属性,通过标准化权重算法计算待融合对象的空间相似度得到融合集;其次,利用非空间Jaro-Winkle算法对融合集中类别一致的对象使用低阈值排除,对类别不一致的使用高阈值排除;最后,使用距离约束、类别一致约束和高阈值的非空间Jaro-Winkle算法找出空间算法遗漏的可融合对象。实验结果表明,该方法平均准确率达到93.3%,与空间和非空间算法（COM-NWT）及格网化纠正方法相比,在7组不同重合度的数据下MNMDC方法的平均准确率提高2.7和1.6个百分点、平均召回率提高2.3和1.4个百分点。MNMDC在实际融合过程中能更精确地融合POI数据。     

Discovering Functional Organized Point of Interest Groups for Spatial Keyword Recommendation

A point of interest (POI) is a specific point location that someone may find useful. With the development of urban modernization, a large number of functional organized POI groups (FOPGs), such as shopping malls, electronic malls, and snacks streets, are springing up in the city. They have a great influence on people’s lives. We aim to discover functional organized POI groups for spatial keyword recommendation because FOPGs-based recommendation is superior to POIs-based recommendation in efficiency and flexibility. To discover FOPGs, we design clustering algorithms to obtain organized POI groups (OPGs) and utilize OPGs-LDA (Latent Dirichlet Allocation) model to reveal functions of OPGs for further recommendation. To the best of our knowledge, we are the first to study functional organized POI groups which have important applications in urban planning and social marketing.     

Delineating urban functional use from points of interest data with neural network embedding: A case study in Greater London

Delineating urban functional use plays a key role in understanding urban dynamics, evaluating planning strategies and supporting policymaking. In recent years, Points of Interest (POI) data, with precise geolocation and detailed attributes, have become the primary data source for exploring urban functional use from a bottom-up perspective, using local, highly disaggregated, big datasets. Previous studies using POI data have given insufficient consideration to the relationship among POI classes in the spatial context, and have failed to provide a straightforward means by which to classify urban functional areas. This study proposes an approach for delineating urban functional use at the scale of the Lower Layer Super Output Area (LSOA) in Greater London by integrating the Doc2Vec model, a neural network embedding method commonly used in natural language processing for vectoring words and documents from their context. In this study, the neural network vectorises both POI classes (‘Word’) and urban areas (‘Document’) based on their functional context by learning features from the spatial distribution of POIs in the city. Specifically, we first construct POI sequences based on the distribution of POI classes, and add their LSOA IDs as ‘document’ tags. By utilising these constructed POI–LSOA sequences, the Doc2Vec model trains the vectors of 574 POI classes (word vectors) and 4836 LSOAs (document vectors). The vectors of POI classes are then used in calculating the functional similarity scores based on their cosine distance, with the vectors of LSOAs grouped into clusters (i.e., functional areas) via the k-means clustering algorithm. We also identify latent functions in each cluster of LSOAs by performing topic modelling and enrichment factor. Compared with TF–IDF, LDA and Word2Vec models, the Doc2Vec model obtains the highest accuracy when classifying functional areas. This study proposes a straightforward approach in which the model directly trains vectors for urban areas, subsequently using them to classify urban functional areas. By employing the enhanced neural network model with low-cost and ubiquitous POI datasets, this study provides a potential tool with which to monitor urban dynamics in a timely and adaptive manner, thereby providing enhanced, data-driven support to urban planning, development and management.     

空间语义增强下的城市交通事故数据可视分析

目的 海量城市交通事故数据可能蕴含有交通事故的空间模式，挖掘出交通事故的空间模式有助于开展交通事故的防治工作。目前交通管理部门虽然记录了交通事故发生地的空间位置信息，但没有对事故发生地进行空间语义描述，从而影响对交通事故空间模式的深入分析。因此，提出一种交通事故数据空间语义增强方法，并设计了一套可视分析系统。方法 基于城市兴趣点来增强交通事故数据的空间语义。以事故发生点为中心获取周围城市兴趣点，使用特征向量刻画兴趣点的数量、类别及其与事故发生点的距离，并称此向量为空间语义特征向量。将空间语义特征向量和相应的交通事故关联，以达到增强其空间语义的目的。然后，基于空间语义特征向量，使用自组织映射聚类算法对交通事故进行聚类分析，根据其空间语义特征将交通事故分为若干类别。最后，通过使用地图视图展示事故点数据、聚类视图和平行坐标视图展示聚类分析的结果及其空间语义特征的可视化方法，对交通事故的空间模式进行分析。结果 针对空间语义增强的交通事故数据以及相关分析任务，有效地使用上述数据分析方法与可视化技术，设计并实现了一套多视图关联的可视分析系统，提供了便捷的交互方式辅助用户分析。通过研发人员和交通警察共同对安徽省合肥市2018年的交通事故数据进行分析，将交通事故发生地划分9类并指出每类地点的空间语义特点，进一步分析出了事故高发区域的空间语义特性。结论 本文提出的交通事故数据空间语义增强方法和可视分析方法可以帮助用户揭示交通事故的空间语义模式，有助于深入分析和认识交通事故的成因，能为交通事故防治相关的城市建设工作提供建议。     

Analyzing urban population data using generalized gamma model and wave-spectrum relation: A case study of Kaohsiung

A quantitative understanding of complex urban growth patterns and processes is crucial to sustainable land management and urban development planning in cities. The spatial organization of urban patterns can be treated as fractals and can be characterized with fractal dimension. However, the calculation of fractal dimension of urban form is often constrained by imperfect and incomplete higher temporal resolution land-use data. Because census data are easily acquired, this study aims to provide a systematic investigation of the relationships between population and urban growth by analyzing changes in urban form that are characterized by fractal dimensions. If the population density in cities follows the negative exponential distribution in proximity, we can use the generalized gamma model and wave-spectrum relation to indirectly estimate the fractal dimension of land-use form in cities. Correlogram analysis is performed to consolidate the results from wave-spectrum relation. Information entropy of the city’s population distribution profile along the radial is calculated to measure the degree of spatial dispersion. The schematic framework is applied to the city of Kaohsiung to get significant insight in the dynamics of pattern formation of the urban population. This is critical for further computer-simulated experiments on urban growth and spatial complexity.     

A web-based pervasive recommendation system for mobile tourist guides

Mobile tourist guides have attracted considerable research interest during the past decade, resulting in numerous standalone and web-based mobile applications. Particular emphasis has been given to personalization of services, typically based on travel recommender systems used to assist tourists in choosing places to visit; these systems address an important aspect of personalization and hence reduce the information burden for the user. However, existing systems fail to exploit information, behaviours, evaluations or ratings of other tourists with similar interests, which would potentially provide ground for the cooperative production of improved tourist content and travel recommendations. In this paper, we extend this notion of travel recommender systems utilizing collaborative filtering techniques while also taking into account contextual information (such as the current user’s location, time, weather conditions and places already visited by the user) for deriving improved recommendations in pervasive environments. We also propose the use of wireless sensor network (WSN) installations around tourist sites for enabling precise localization and also providing mobile users convenient and inexpensive means for uploading tourist information and ratings about points of interest (POI) via their mobile devices. We also introduce the concept of ‘context-aware rating’, whereby user ratings uploaded through WSN infrastructures are weighted higher to differentiate among users that rate POIs using the mobile tourist guide application while onsite and others using the Internet away from the POI.     

Metro Transit‐Centric Visualization for City Tour Planning

In general, city trip planning consists of two main steps: knowing Points‐Of‐Interest (POIs), and then planning a tour route from the current point to next preferred POIs. We mainly consider the metro for traveling around touristic cities as the main means of transportation. In this context, existing tools lack a capability to effectively visualize POIs on the metro map for trip planning. To bridge this gap, we propose an interactive framework that holistically combines presentations of POIs and a metro network. Our idea is to identify popular POIs based on visual worth computation, and to introduce POI discovery for effectively identifying POIs within reach of a metro network for users. We use octilinear layouts to highlight the metro network, and show representative POI images in the layout space visualized within a user‐specified viewing window. We have implemented our working prototype showing touristic cities with a metro network. We have factored out various design guidelines that are basis for designing our method, and validated our approach with a user study surveying 70 individuals.     

How do you visit: Identifying addicts from large-scale transit records via scenario deep embedding

Identification of individuals based on transit modes is of great importance in user tracking systems. However, identifying users in real-life studies is not trivial owing to the following challenges: 1) activity data containing both temporal and spatial context are high-order and sparse; 2) traditional two-step classifiers depend on trajectory patterns as input features, which limits accuracy especially in the case of scattered and diverse data; 3) in some cases, there are few positive instances and they are difficult to detect. Therefore, approaches involving statistics-based or trajectory-based features do not work effectively. Deep learning methods also suffer from the problem of how to represent trajectory vectors for user classification. Here, we propose a novel end-to-end scenario-based deep learning method to address these challenges, based on the observation that individuals may visit the same place for different reasons. We first define a scenario using critical places and related trajectories. Next, we embed scenarios via path-based or graph-based approaches using extended embedding techniques. Finally, a two-level convolution neural network is constructed for the classification. Our model is applied to the problem of detection of addicts using transit records directly without feature engineering, based on real-life data collected from mobile devices. Based on constructed scenario with dense trajectories, our model outperforms classical classification approaches, anomaly detection methods, state-of-the-art sequential deep learning models, and graph neural networks. Moreover, we provide statistical analyses and intuitiveexplanations to interpret the characteristics of resident and addict mobility. Our method could be generalized to other trajectory-related tasks involving scattered and diverse data.