A graph-based approach to detecting tourist movement patterns using social media data

ABSTRACT

Understanding the characteristics of tourist movement is essential for tourist behavior studies since the characteristics underpin how the tourist industry management selects strategies for attraction planning to commercial product development. However, conventional tourism research methods are not either scalable or cost-efficient to discover underlying movement patterns due to the massive datasets. With advances in information and communication technology, social media platforms provide big data sets generated by millions of people from different countries, all of which can be harvested cost efficiently. This paper introduces a graph-based method to detect tourist movement patterns from Twitter data. First, collected tweets with geo-tags are cleaned to filter those not published by tourists. Second, a DBSCAN-based clustering method is adapted to construct tourist graphs consisting of the tourist attraction vertices and edges. Third, network analytical methods (e.g. betweenness centrality, Markov clustering algorithm) are applied to detect tourist movement patterns, including popular attractions, centric attractions, and popular tour routes. New York City in the United States is selected to demonstrate the utility of the proposed methodology. The detected tourist movement patterns assist business and government activities whose mission is tour product planning, transportation, and development of both shopping and accommodation centers.     

A spatial‐knowledge‐enhanced heuristic for solving the p‐median problem

The p‐median problem (PMP) is one of the most applied location problems in urban and regional planning. As an NP‐hard problem, the PMP remains challenging to solve optimally, especially for large‐sized problems. A number of heuristics have been developed to obtain PMP solutions in a fast manner. Among the heuristics, the Teitz and Bart (TB) algorithm has been found effective for finding high‐quality solutions. In this article, we present a spatial‐knowledge‐enhanced Teitz and Bart (STB) heuristic method for solving PMPs. The STB heuristic prioritizes candidate facility sites to be examined in the solution set based on the spatial distribution of demand and service provision. Tests based on a range of PMPs demonstrate the effectiveness of the STB heuristic. This new algorithm can be incorporated into current commercial GIS packages to solve a wide range of location‐allocation problems.     

Building a model-based personalised recommendation approach for tourist attractions from geotagged social media data

When travelling, people are accustomed to taking and uploading photos on social media websites, which has led to the accumulation of huge numbers of geotagged photos. Combined with multisource information (e.g. weather, transportation, or textual information), these geotagged photos could help us in constructing user preference profiles at a high level of detail. Therefore, using these geotagged photos, we built a personalised recommendation system to provide attraction recommendations that match a user's preferences. Specifically, we retrieved a geotagged photo collection from the public API for Flickr (Flickr.com) and fetched a large amount of other contextual information to rebuild a user's travel history. We then created a model-based recommendation method with a two-stage architecture that consists of candidate generation (the matching process) and candidate ranking. In the matching process, we used a support vector machine model that was modified for multiclass classification to generate the candidate list. In addition, we used a gradient boosting regression tree to score each candidate and rerank the list. Finally, we evaluated our recommendation results with respect to accuracy and ranking ability. Compared with widely used memory-based methods, our proposed method performs significantly better in the cold-start situation and when mining ‘long-tail’ data.     

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。     

Measuring urban sentiments from social media data: a dual-polarity metric approach

Journal of Geographical Systems - Urban sentiment, as people’ perception of city environment and events, is a direct indicator of the quality of life of residents and the unique identity of a...     

Mapping fine‐scale urban housing prices by fusing remotely sensed imagery and social media data

The accurate mapping of urban housing prices at a fine scale is essential to policymaking and urban studies, such as adjusting economic factors and determining reasonable levels of residential subsidies. Previous studies focus mainly on housing price analysis at a macro scale, without fine‐scale study due to a lack of available data and effective models. By integrating a convolutional neural network for united mining (UMCNN) and random forest (RF), this study proposes an effective deep‐learning‐based framework for fusing multi‐source geospatial data, including high spatial resolution (HSR) remotely sensed imagery and several types of social media data, and maps urban housing prices at a very fine scale. With the collected housing price data from China's biggest online real estate market, we produced the spatial distribution of housing prices at a spatial resolution of 5 m in Shenzhen, China. By comparing with eight other multi‐source data mining techniques, the UMCNN obtained the highest housing price simulation accuracy (Pearson R = 0.922, OA = 85.82%). The results also demonstrated a complex spatial heterogeneity inside Shenzhen's housing price distribution. In future studies, we will work continuously on housing price policymaking and residential issues by including additional sources of spatial data.     

NeuroTPR: A neuro‐net toponym recognition model for extracting locations from social media messages

Social media messages, such as tweets, are frequently used by people during natural disasters to share real‐time information and to report incidents. Within these messages, geographic locations are often described. Accurate recognition and geolocation of these locations are critical for reaching those in need. This article focuses on the first part of this process, namely recognizing locations from social media messages. While general named entity recognition tools are often used to recognize locations, their performance is limited due to the various language irregularities associated with social media text, such as informal sentence structures, inconsistent letter cases, name abbreviations, and misspellings. We present NeuroTPR, which is a Neuro‐net ToPonym Recognition model designed specifically with these linguistic irregularities in mind. Our approach extends a general bidirectional recurrent neural network model with a number of features designed to address the task of location recognition in social media messages. We also propose an automatic workflow for generating annotated data sets from Wikipedia articles for training toponym recognition models. We demonstrate NeuroTPR by applying it to three test data sets, including a Twitter data set from Hurricane Harvey, and comparing its performance with those of six baseline models.     

Optimal spatial data matching for conflation: A network flow‐based approach

Spatial data conflation involves the matching and merging of counterpart features in multiple datasets. It has applications in practical spatial analysis in a variety of fields. Conceptually, the feature‐matching problem can be viewed as an optimization problem of seeking a match plan that minimizes the total discrepancy between datasets. In this article, we propose a powerful yet efficient optimization model for feature matching based on the classic network flow problem in operations research. We begin with a review of the existing optimization‐based methods and point out limitations of current models. We then demonstrate how to utilize the structure of the network‐flow model to approach the feature‐matching problem, as well as the important factors for designing optimization‐based conflation models. The proposed model can be solved by general linear programming solvers or network flow solvers. Due to the network flow formulation we adopt, the proposed model can be solved in polynomial time. Computational experiments show that the proposed model significantly outperforms existing optimization‐based conflation models. We conclude with a summary of findings and point out directions of future research.     

A spectral-geophysical approach for detecting pipeline leakage

Leakage of hydrocarbon has a large economic and environmental impact. Traditional methods for investigating leakage and resulting pollution, such as drilling, are destructive, time consuming and expensive. Remote sensing is an alternative that is non-destructive and has been been tested extensively for exploration of onshore hydrocarbon reservoirs and detection of hydrocarbons at the Earth’s surface. In this research, a leaking pipeline is investigated through field reflectance spectrometry and the findings are validated with traditional drilling and geophysical measurements. The measurements show a significant increase of vegetation anomalies on the pipeline with respect to areas further away. The observed anomalies are positively related to hydrocarbon pollution through chemical analysis of drillings. Subsurface geophysical measurements show a large correlation with observed surface vegetation stress, enhancing the identification of hydrocarbon-related vegetation stress through spectroscopy.     

Enhancing data privacy with semantic trajectories: A raster‐based framework for GPS stop/move management

Tracking facilities on smartphones generate enormous amounts of GPS trajectories, which provide new opportunities to study movement patterns and improve transportation planning. Converting GPS trajectories into semantically meaningful trips is attracting increasing research effort with respect to the development of algorithms, frameworks, and software tools. There are, however, few works focused on designing new semantic enrichment functionalities taking privacy into account. This article presents a raster‐based framework which not only detects significant stop locations, segments GPS records into stop/move structures, and brings semantic insights to trips, but also provides possibilities to anonymize users’ movements and sensitive stay/move locations into raster cells/regions so that a multi‐level data sharing structure is achieved for a variety of data sharing purposes.     

Spatio‐temporal prediction of land surface temperature using semantic kriging

Spatio‐temporal prediction and forecasting of land surface temperature (LST) are relevant. However, several factors limit their usage, such as missing pixels, line drops, and cloud cover in satellite images. Being measured close to the Earth's surface, LST is mainly influenced by the land use/land cover (LULC) distribution of the terrain. This article presents a spatio‐temporal interpolation method which semantically models LULC information for the analysis of LST. The proposed spatio‐temporal semantic kriging (ST‐SemK) approach is presented in two variants: non‐separable ST‐SemK (ST‐SemK_NSep) and separable ST‐SemK (ST‐SemK_Sep). Empirical studies have been carried out with derived Landsat 7 ETM+ satellite images of LST for two spatial regions: Kolkata, India and Dallas, Texas, U.S. It has been observed that semantically enhanced spatio‐temporal modeling by ST‐SemK yields more accurate prediction results than spatio‐temporal ordinary kriging and other existing methods.     

A heuristic‐based approach to mitigating positional errors in patrol data for species distribution modeling

Species distribution modeling (SDM) at fine spatial resolutions requires species occurrence data of high positional accuracy to achieve good model performance. However, wildlife occurrences recorded by patrols in ranger‐based monitoring programs suffer from positional errors, because recorded locations represent the positions of the ranger and differ from the actual occurrence locations of wildlife (hereinafter referred to as positional errors in patrol data). This study presented an evaluation of the impact of such positional errors in patrol data on SDM and developed a heuristic‐based approach to mitigating the positional errors. The approach derives probable wildlife occurrence locations from ranger positions, utilizing heuristics based on species preferred habitat and the observer's field of view. The evaluations were conducted through a case study of SDM using patrol records of the black‐and‐white snub‐nosed monkey (Rhinopithecus bieti) in Yunnan, China. The performance of the approach was also compared against alternative sampling methods. The results showed that the positional errors in R. bieti patrol data had an adverse effect on SDM performance, and that the proposed approach can effectively mitigate the impact of the positional errors to greatly improve SDM performance.     

局部表面参数化的实景三角网模型语义增强方法

倾斜摄影测量方法已可自动获取城市规模的实景三角网模型,然而散乱的三角网缺乏精细的几何结构和功能语义信息。为克服上述问题,提出一种局部表面参数化的实景三角网模型语义增强方法:将具有语义信息的独立三维部件与实景三角网模型的无缝融合问题,通过定义三维表面结构树,转换为局部区域的三角网替换操作;在待融合区域附近,将原实景三角网模型和替换的三维语义部件,通过局部参数化表达,UV展开为二维平面三角网;在二维平面上构建约束Delaunay三角网(CDT),实现两模型的无缝拼接,逆映射至三维空间并自动重建语义部件。通过深圳某区域的倾斜影像进行的试验证明,本文方法能有效实现具有开放边界和语义信息的部件模型与表面模型的无缝融合。与商业软件Maya对比,这种基于插入、融合的手段对提高建模效率具有实用价值。     

Latent spatio-temporal activity structures: a new approach to inferring intra-urban functional regions via social media check-in data

Abstract

This article introduces a novel low rank approximation (LRA)-based model to detect the functional regions with the data from about 15 million social media check-in records during a year-long period in Shanghai, China. We identified a series of latent structures, named latent spatio-temporal activity structures. While interpreting these structures, we can obtain a series of underlying associations between the spatial and temporal activity patterns. Moreover, we can not only reproduce the observed data with a lower dimensional representative, but also project spatio-temporal activity patterns in the same coordinate system. With the K-means clustering algorithm, five significant types of clusters that are directly annotated with a combination of temporal activities can be obtained, providing a clear picture of the correlation between the groups of regions and different activities at different times during a day. Besides the commercial and transportation dominant areas, we also detected two kinds of residential areas, the developed residential areas and the developing residential areas. We further interpret the spatial distribution of these clusters using urban form analytics. The results are highly consistent with the government planning in the same periods, indicating that our model is applicable to infer the functional regions from social media check-in data and can benefit a wide range of fields, such as urban planning, public services, and location-based recommender systems.     

Extraction of spatio‐temporal data about historical events from text documents

Often, we are faced with questions regarding past events and the answers are hidden in the historical text archives. The growing developments in geographic information retrieval and temporal information retrieval techniques have given new ways to explore digital text archives for spatio‐temporal data. The question is how to retrieve the answers from the text documents. This work contributes to a better understanding of spatio‐temporal information extraction from text documents. Natural language processing techniques were used to develop an information extraction approach using the GATE language processing software. The developed framework uses gazetteer matching, spatio‐temporal relationship extraction and pattern‐based rules to recognize and annotate elements in historical text documents. The extracted spatio‐temporal data is used as input for GIS studies on the time–geography context of the German–Herero resistance war of 1904 in Namibia. Related issues when analyzing the historical data in current GIS are discussed. Particularly problematic are movement data in small scale with poor temporal density and trajectories that are short or connect very distant locations.     

A name‐led approach to profile urban places based on geotagged Twitter data

Place is a concept that is fundamental to how we orientate and communicate space in our everyday lives. Crowdsourced social media data present a valuable opportunity to develop bottom‐up inferences of places that are integral to social activities and settings. Conventional location‐led approaches use a predefined spatial unit to associate data and space with places, which cannot capture the richness of urban places (i.e., spatial extents and their dynamic functions). This article develops a name‐led framework to overcome these limitations in using social media data to study urban places. The framework first derives place names from georeferenced Twitter data combining text mining and spatial point pattern analysis, then estimates the spatial extents by spatial clustering, and further extracts their dynamic functions with time, which makes up a complete place profile. The framework is tested on a case study in Camden Borough, London and the results are evaluated through comparisons to the Foursquare point of interest data. This name‐led approach enables the shift from space‐based analysis to place‐based analysis of urban space.     

On the feasibility of using satellite gravity observations for detecting large-scale solid mass transfer events

The main focus of this paper is to assess the feasibility of utilizing dedicated satellite gravity missions in order to detect large-scale solid mass transfer events (e.g. landslides). Specifically, a sensitivity analysis of Gravity Recovery and Climate Experiment (GRACE) gravity field solutions in conjunction with simulated case studies is employed to predict gravity changes due to past subaerial and submarine mass transfer events, namely the Agulhas slump in southeastern Africa and the Heart Mountain Landslide in northwestern Wyoming. The detectability of these events is evaluated by taking into account the expected noise level in the GRACE gravity field solutions and simulating their impact on the gravity field through forward modelling of the mass transfer. The spectral content of the estimated gravity changes induced by a simulated large-scale landslide event is estimated for the known spatial resolution of the GRACE observations using wavelet multiresolution analysis. The results indicate that both the Agulhas slump and the Heart Mountain Landslide could have been detected by GRACE, resulting in \({\vert }0.4{\vert }\) and \({\vert }0.18{\vert }\) mGal change on GRACE solutions, respectively. The suggested methodology is further extended to the case studies of the submarine landslide in Tohoku, Japan, and the Grand Banks landslide in Newfoundland, Canada. The detectability of these events using GRACE solutions is assessed through their impact on the gravity field.     

Analyzing relationship between user‐generated content and local visual information with augmented reality‐based location‐based social networks

Location‐based social networks (LBSNs) have become an important source of spatial data for geographers and GIScientists to acquire knowledge of human–place interactions. A number of studies have used geotagged data from LBSNs to investigate how user‐generated content (UGC) can be affected by or correlated with the external environment. However, local visual information at the micro‐level, such as brightness, colorfulness, or particular objects/events in the surrounding environment, is usually not captured and thus becomes a missing component in LBSN analysis. To provide a solution to this issue, we argue in this study that the integration of augmented reality (AR) and LBSNs proves to be a promising avenue. In this first empirical study on AR‐based LBSNs, we propose a methodological framework to extract and analyze data from AR‐based LBSNs and demonstrate the framework via a case study with WallaMe. Our findings bolster existing psychological findings on the color–mood relationship and display intriguing geographic patterns of the influence of local visual information on UGC in social media.     

A multi‐criteria decision‐making approach for geometric matching of areal objects

We propose a method for geometric areal object matching based on multi‐criteria decision making. To enable this method, we focused on determining the matched areal object pairs that have all relations, one‐to‐one relationships to many‐to‐many relationships, in different spatial data sets by fusing geometric criteria without user invention. First, we identified candidate corresponding areal object pairs with a graph‐based approach in training data. Second, three matching criteria (areal hausdorff distance, intersection ratio, and turning function distance) were calculated in candidate corresponding pairs and these criteria were normalized. Third, the shape similarity was calculated by weighted linear combination using the normalized matching criteria (similarities) with the criteria importance through intercriteria correlation method. Fourth, a threshold (0.738) of the shape similarity estimated in the plot of precision versus recall versus all possible thresholds of training data was applied, and the matched pairs were determined and identified. Finally, we visually validated the detection of similar areal feature pairs and conducted statistical evaluation using precision, recall, and F‐measure values from a confusion matrix. Their values were 0.905, 0.848, and 0.876, respectively. These results validate that the proposed classifier, which detects 87.6% of matched areal pairs, is highly accurate.     

A data-driven approach to local gravity field modelling using spherical radial basis functions

We propose a methodology for local gravity field modelling from gravity data using spherical radial basis functions. The methodology comprises two steps: in step 1, gravity data (gravity anomalies and/or gravity disturbances) are used to estimate the disturbing potential using least-squares techniques. The latter is represented as a linear combination of spherical radial basis functions (SRBFs). A data-adaptive strategy is used to select the optimal number, location, and depths of the SRBFs using generalized cross validation. Variance component estimation is used to determine the optimal regularization parameter and to properly weight the different data sets. In the second step, the gravimetric height anomalies are combined with observed differences between global positioning system (GPS) ellipsoidal heights and normal heights. The data combination is written as the solution of a Cauchy boundary-value problem for the Laplace equation. This allows removal of the non-uniqueness of the problem of local gravity field modelling from terrestrial gravity data. At the same time, existing systematic distortions in the gravimetric and geometric height anomalies are also absorbed into the combination. The approach is used to compute a height reference surface for the Netherlands. The solution is compared with NLGEO2004, the official Dutch height reference surface, which has been computed using the same data but a Stokes-based approach with kernel modification and a geometric six-parameter “corrector surface” to fit the gravimetric solution to the GPS-levelling points. A direct comparison of both height reference surfaces shows an RMS difference of 0.6 cm; the maximum difference is 2.1 cm. A test at independent GPS-levelling control points, confirms that our solution is in no way inferior to NLGEO2004.