JiuTian·Chuanliu: A Large Spatiotemporal Model for General-purpose Dynamic Urban Sensing

The impacts of disasters are increasing due to climate change and unplanned urbanization. Big and open data offer considerable potential for analyzing and predicting human mobility during disaster events, including the COVID-19 pandemic, leading to better disaster risk reduction (DRR) planning. However, the value of human mobility data and analysis (HMDA) in urban resilience research is poorly understood. This review highlights key opportunities for and challenges hindering the use of HMDA in DRR in urban planning and risk science, as well as insights from practitioners. A gap in research on HMDA for data-driven DRR planning was identified. By examining human mobility studies and their respective analytical and planning tools, this paper offers deeper insights into the challenges that must be addressed to improve the development of effective data-driven DRR planning, from data collection to implementation. In future work on HMDA, (i) the human mobility of vulnerable populations should be targeted, (ii) research should focus on disaster mitigation and prevention, (iii) analytical methods for evidence-based disaster planning should be developed, (iv) different types of data should be integrated into analyses to overcome methodological challenges, and (v) a decision-making framework should be developed for evidence-based urban planning through transdisciplinary knowledge co-production.

With the development of sensors, recording and availability of high-resolution movement data from animals and humans, two disciplines have rapidly developed: human mobility and movement ecology. Addressing methodological gaps between these two mobility fields could improve the understanding of movement processes and has been defined as the Integrated Science of Movement. We apply well-known human mobility metrics and data processing methods to Global Positioning System (GPS) tracking data of European Herring Gulls (Larus argentatus) to test the usefulness of these methods for explaining animal mobility behavior. We use stop detection, spatial aggregation, and for the first time on animal movement data, two approaches to temporal aggregation (Next Time-Bin and Next Place). We also calculate from this data a set of movement statistics (visitation frequency, distinct locations over time, and radius of gyration). Furthermore, we analyze and compare the gull and human data from the perspective of scaling laws commonly used for human mobility. The results confirm those of previous studies and indicate differences in movement parameters between the breeding season and other parts of the year. This paper also shows that methods used in human mobility analysis have the potential to improve our understanding of animal behavior.

The use of virtual reality games, known as “exergaming”, is gaining more and more interest as a mobilization tool and as a key piece in the delivery of quality health, especially in elderly people. Mobility tracking of elderly people facilitates the extraction of useful spatiotemporal characteristics regarding their activities and behavior at home. Currently, the analysis of human mobility is based on expensive technologies. In this paper, we propose a semantic interoperability agent which exploits mobility tracking and spatiotemporal characteristics to extract human profiling and give incentives for mobilization at home. The agent exploits an extended ontology which facilitates the collation of evidence for the effects of exergaming on the movement control of older adults. In order to provide personalized monitoring services, a number of rules are individually defined to generate incentives. To evaluate the proposed semantic interoperability agent, human mobility data are collected and analyzed based on daily activities, their duration and mobility patterns. We show that the proposed agent is robust enough for activity classification, and that the recommendations for mobilization are accurate. We further demonstrate the agent's potential in useful knowledge inference regarding personalized elderly people home care.

Advances in sensor, wireless communication, and information infrastructure such as GPS have enabled us to collect massive amounts of human mobility data, which are fine-grained and have global road coverage. These human mobility data, if properly encoded with semantic information (i.e. combined with Point of Interests (POIs)), is appealing for changing the paradigm for gas station site selection. To this end, in this paper, we investigate how to exploit newly-generated human mobility data for enhancing gas station selection. Specifically, we develop a ranking system for evaluating the business performances of gas stations based on waiting time of refueling events by mining human mobility data. Along this line, we first design a method for detecting taxi refueling events by jointly tracking dwell times, GPS trace angles, location sequences, and refueling cycles of the vehicles. Also, we extract the fine-grained discriminative features strategically from POI data, human mobility data and road network data within the neighborhood of gas stations, and perform feature selection by simultaneously maximizing relevance and minimizing redundancy based on mutual information. In addition, we learn a ranking model for predicting gas station crowdedness by exploiting learning to rank techniques. The extensive experimental evaluation on real-world data also show the advantages of the proposed method over existing approaches for gas site selection.

ABSTRACTWater demand forecasting is a crucial task in the efficient management of the water supply system. This paper compares classical and adapted machine learning algorithms used for water usage predictions including ARIMA, support vector regression, random forests and extremely randomized trees. These models were enriched with human mobility data to improve the predictive power of water demand forecasting. Furthermore, a framework for processing mobility data into time-series correlated with water usage data is proposed. This study uses 51 days of water consumption readings and over 7 million geolocated mobility records from urban areas. Results show that using human mobility data improves water demand prediction. The best forecasting algorithm employing a random forest method achieved 90.4% accuracy (measured by the mean absolute percentage error) and is better by 1% than the same algorithm using only water data, while classic ARIMA approach achieved 90.0%. The Blind (copying) prediction achieved 85.1% of accuracy.

Epidemic modeling and simulation techniques have been getting attention since the outbreak of Ebola virus in West Africa regions in 2014. The epidemic modeling has been significantly studied so far, but recent improvements of high performance computing technologies and availability of human mobility and contact data promoted by the IoT enables more realistic data-driven approaches. In this paper, we introduce the agent-based infections diffusion simulation using real human mobility data as metapopulation network. We compare the results of simulations with the conventional measures of network analysis, which highlights the metapopulation networks have important roles of pandemic in society.

The COVID-19 pandemic poses unprecedented challenges around the world. Many studies indicate that human mobility data provide significant support for public health actions during the pandemic. Researchers have applied mobility data to explore spatiotemporal trends over time, investigate associations with other variables, and predict or simulate the spread of COVID-19. Our objective was to provide a comprehensive overview of human mobility open data to guide researchers and policymakers in conducting data-driven evaluations and decision-making for the COVID-19 pandemic and other infectious disease outbreaks. We summarized the mobility data usage in COVID-19 studies by reviewing recent publications on COVID-19 and human mobility from a data-oriented perspective. We identified three major sources of mobility data: public transit systems, mobile operators, and mobile phone applications. Four approaches have been commonly used to estimate human mobility: public transit-based flow, social activity patterns, index-based mobility data, and social media-derived mobility data. We compared mobility datasets’ characteristics by assessing data privacy, quality, space-time coverage, high-performance data storage and processing, and accessibility. We also present challenges and future directions of using mobility data. This review makes a pivotal contribution to understanding the use of and access to human mobility data in the COVID-19 pandemic and future disease outbreaks.

Human mobility data are an important proxy to understand human mobility dynamics, develop analytical services, and design mathematical models for simulation and what-if analysis. Unfortunately mobility data are very sensitive since they may enable the re-identification of individuals in a database. Existing frameworks for privacy risk assessment provide data providers with tools to control and mitigate privacy risks, but they suffer two main shortcomings: (i) they have a high computational complexity; (ii) the privacy risk must be recomputed every time new data records become available and for every selection of individuals, geographic areas, or time windows. In this article, we propose a fast and flexible approach to estimate privacy risk in human mobility data. The idea is to train classifiers to capture the relation between individual mobility patterns and the level of privacy risk of individuals. We show the effectiveness of our approach by an extensive experiment on real-world GPS data in two urban areas and investigate the relations between human mobility patterns and the privacy risk of individuals.

Ranking locations in a city via the collective home-work relations in human mobility data

Human mobility patterns reflect many aspects of life, from the global spread of infectious diseases to urban planning and daily commute patterns. In recent years, the prevalence of positioning methods and technologies, such as the global positioning system, cellular radio tower geo-positioning, and WiFi positioning systems, has driven efforts to collect human mobility data and to mine patterns of interest within these data in order to promote the development of location-based services and applications. The efforts to mine significant patterns within large-scale, high-dimensional mobility data have solicited use of advanced analysis techniques, usually based on machine learning methods, and therefore, in this paper, we survey and assess different approaches and models that analyze and learn human mobility patterns using mainly machine learning methods. We categorize these approaches and models in a taxonomy based on their positioning characteristics, the scale of analysis, the properties of the modeling approach, and the class of applications they can serve. We find that these applications can be categorized into three classes: user modeling, place modeling, and trajectory modeling, each class with its characteristics. Finally, we analyze the short-term trends and future challenges of human mobility analysis.

Human mobility nowcasting is a fundamental research problem for intelligent transportation planning, disaster responses and management, etc. In particular, human mobility under big disasters such as hurricanes and pandemics deviates from its daily routine to a large extent, which makes the task more challenging. Existing works mainly focus on traffic or crowd flow prediction in normal situations. To tackle this problem, in this study, disaster-related Twitter data is incorporated as a covariate to understand the public awareness and attention about the disaster events and thus perceive their impacts on the human mobility. Accordingly, we propose a Meta-knowledge-Memorizable Spatio-Temporal Network (MemeSTN), which leverages memory network and meta-learning to fuse social media and human mobility data. Extensive experiments over three real-world disasters including Japan 2019 typhoon season, Japan 2020 COVID-19 pandemic, and US 2019 hurricane season were conducted to illustrate the effectiveness of our proposed solution. Compared to the state-of-the-art spatio-temporal deep models and multivariate-time-series deep models, our model can achieve superior performance for nowcasting human mobility in disaster situations at both country level and state level.

In this paper, we investigate the problem of detecting real-time city-scale hyper-local events based on the analysis of social media and human mobility. Different from general events reported by news media, hyper-local events refer to both small-scale and large-scale events pertaining to a geographical location. Since small-scale events, e.g, a party in a pub, an exhibition in a local museum, are not often reported through mainstream media platforms, such as newspapers, TV news, web media or government report, it is challenging to obtain the resources related to such kind of events. Besides, those media platforms have a great latency in reporting the news of ongoing events, resulting in that the events we saw might take place a couple of days ago. Though people have tried to find clues of events from real-time social media streams (e.g., Instagram and Twitter), the scarcity of social posts with geo-tagged information leads to a very low quality of localized event detection. In this paper, in addition to the data from social media stream, we apply human mobility data which contain rich spatial-temporal information as another important resource to improve the performance of hyper-local event detection. Specifically, we use taxi data as it is expected that the occurrence of hyper-local events usually leads to the change in the surrounding traffic. As far as our knowledge, this is the first work which combines multiple social media data sources with human mobility information for the task of real-time hyper-local event detection. We propose a two-step framework which is composed of an anomaly filter and an event classifier. Through experiments on New York City data, we show that our proposed system can effectively detect both small-scale and large-scale local events. Furthermore, we verify that applying human mobility data can significantly enhance the performance of event detection and classification.

Human mobility is central to our understanding of design, planning and development of civil infrastructure, particularly in urban areas where large scale mobility flow problems can critically depend on the interface between human mobility and infrastructure. Therefore, researchers have spent considerable effort to understand and predict human mobility patterns. Several recent studies have used geo-social networking platforms to examine human mobility, but the focus of these studies has been on small scale social networking media. In this study, we examined the possibility of using Twitter, a massive online social networking platform with over 400 million users, to collect human mobility data. We developed a process map to collect data from Twitter, and designed two Python modules for its implementation. A case study was conducted and its results confirmed that Twitter can provide a larger quantity of useful human mobility data. In future research, we plan to analyze the data and validate that it can accurately capture mobility patterns. This will provide insight into whether Twitter is a viable resource to study city-scale human mobility. It can also potentially deepen our understanding about the interaction between urban dwellers and civil infrastructure.

Auditing the fairness of place-based crime prediction models implemented with deep learning approaches

Mediation by human mobility of the association between temperature and COVID-19 transmission rate