Towards resilient and smart cities: A real-time urban analytical and geo-visual system for social media streaming data
Introduction
Cities are complex and dynamic systems that contain infrastructures, information, and innovation and house the majority of the world’s population (Batty, 2008). Cities are also exposed to a variety of unforeseeable extreme events, such as disasters and infectious diseases, which have at times caused tremendous economic and social losses (Arafah & Winarso, 2017; Zhu, Li, & Feng, 2019). Disasters have affected more than a third of the world’s population (1.5 billion) and cost more than US$1.3 trillion in economic losses (UN DESA Population Division, 2018). In recent years, influenza epidemics have caused up to 56,000 deaths annually in the United States and have had substantial financial costs (McGowan et al., 2019). To respond actively to such events, researchers and practitioners from multiple disciplines develop theories and approaches to help cities prepare for unexpected perturbations (Woetzel et al., 2018; Zhang & Li, 2018).
In light of the need to building resilient and smart cities in this context, urban studies and practices strive to maintain cities’ essential functionality while reducing the adverse effects when disruptions happen (Allam & Newman, 2018; Angelidou et al., 2018; Desouza & Flanery, 2013; Hatuka, Rosen-Zvi, Birnhack, Toch, & Zur, 2018; Leichenko, 2011; Wang, Hulse, Von Meding, Brown, & Dedenbach, 2019). Existing literature body has discussed four critical aspects of resilience: robustness (the ability to withstand stress without suffering degradation or loss of function), redundancy (the extent to which components can be substituted for to recover reduced or lost functionality), resourcefulness (the capacity to identify problems, establish priorities, and allocate resources), and rapidity (the ability to meet priorities and achieve goals promptly) (Bruneau et al., 2003; Godschalk, 2003; Zobel, 2011). However, as cities grow and gain complexity, conventional approaches that treat resilience as a conceptual process and use static data can become ineffective for achieving the conditions outlined above (Meerow, Newell, & Stults, 2016). Thus, it is necessary to implement new methods and technologies to address the challenges of extreme events and promote resilience.
In the context of burgeoning big data and advanced information and communication technologies (ICTs), more “smart” solutions have also been proposed to help cities survive and function under extreme stresses (Palmieri, Ficco, Pardi, & Castiglione, 2016; Soyata, Habibzadeh, Ekenna, Nussbaum, & Lozano, 2019; Yang, Su, & Chen, 2017). A recent article proposed the smart robustness, smart redundancy, smart resourcefulness, and smart rapidity to leverage resilience by embedding smart technologies and systems in the fabric of cities (Desroches & Taylor, 2018). Although rapidity (e.g., speed) in responding to risks is essential to resilient cities (Al Nuaimi, Al Neyadi, Mohamed, & Al-Jaroodi, 2015; Desouza & Flanery, 2013; Palmieri et al., 2016; Platt, Brown, & Hughes, 2016), few studies have focused on this dimension of urban resilience, particularly among urban-scale quantitative studies (Meerow et al., 2016). Within the smart city context, however, rapid or real-time big data applications can mitigate damages’ impacts and enhance the capacity to recover from extreme events quickly (Desroches & Taylor, 2018; Malik, Sam, Hussain, & Abuarqoub, 2018). For instance, early detection of crises or emergencies and rapid responses allow cities to collect relevant information, monitor the characteristics of events (e.g., locations, time, types), and provide timely analyses and predictions, and thus to better coordinate relief efforts, assess damages, and restore urban system performance (Desouza & Flanery, 2013; Khan, Anjum, Soomro, & Tahir, 2015; Kitchin, 2014; Kontokosta & Malik, 2018; Woetzel et al., 2018; Zhang, Li, Li, & Fang, 2019).
However, most existing quantitative studies are conceptual rather than operational to enhance resilience with smart rapidity, because it is challenging to design a specific plan for an abstract and complex notion such as resilience (Desouza & Flanery, 2013; Hatuka et al., 2018; Wang, Taylor, & Garvin, 2020). For example, Klein, Koenig, and Schmitt (2017) described a vision and a conceptual framework for monitoring and managing cities’ environmental and social dynamics without giving specific methods or plans. Current efforts to create smart and resilient cities also suffer from a mismatch between real-time information resources and delayed decision-making, as well as incompatible algorithms for processing high-volume and -velocity urban streaming data (Al Nuaimi et al., 2015; Khan et al., 2015; Yang et al., 2017). The existing prototypes of urban analytics systems (e.g., Huang, Cervone, & Zhang, 2017; Psyllidis, Bozzon, Bocconi, & Titos Bolivar, 2015) were designed for the analysis and visualization of a diversity of urban topics (human movement patterns, traffic conditions, or place of interests) using periodically updated data or a mixture of static and streaming data. These prototypes did not take full advantage of urban streaming data for smart and rapid resilient city management.
In this research, we propose a real-time urban analytical and visual system that can detect, track, analyze, and visualize small-scale, undefined extreme events clustered in content and space. The system is built on the latest versions of GeoEvent Server and Online GIS for real-time data analysis and visualization and uses geotagged streaming Twitter data. We construct several data-mining and natural language–processing modules within the system, including online topic modeling and sentiment analysis using Apache Spark. The Apache Spark distributed system is especially favorable for online big-data processing with high speed and accuracy. The system is designed for geo-textual streaming data and has the potential to be applied to various urban management scenarios. By leveraging high-volume urban streaming data and smart technologies, we hope to demonstrate the usefulness of our system to understand the dynamics of urban systems, especially during unpredicted perturbations such as disasters. The system demonstrates the analysis results with interactive maps to improve situational awareness and enhance community engagement during extreme events. The system can also be integrated into a holistic, intelligent system to play an active role in future urban planning to achieve smart and resilient cities.
Section snippets
Related work
Recently, extracting and interpreting information from streaming data has gained increasing prominence in the data mining domain. In addition, social media platforms, such as Twitter, have brought valuable user-generated behavior-rich data resources in real time, offering a growing number of opportunities to analyze the dynamics of the text streams and topics (Benhardus & Kalita, 2013; Ghani, Hamid, Hashem, & Ahmed, 2019). In urban contexts, these platforms allow people to share the events they
Developing a real-time urban analytical and geo-visual system
In this section, we demonstrate the design and methods of our real-time urban analytical and geo-visual system that works for streaming geo-textual data (i.e., data with both geographical and content features). This system is built upon Esri’s GeoEvent Server (version 10.7). The server provides comprehensive tools and pipelines to support high-volume real-time data input, processing, and output, making it especially suitable for geospatial streaming data. We employ several tools provided by the
Study case and system settings
We applied the proposed visual urban analytical system in a simulated real-time scenario of Hurricane Harvey, one of the most destructive disasters to happen in the U.S. in the past decade. The hurricane caused more than a hundred billion in damage and made landfall in a densely populated area in south-central Texas. We use geotagged tweets collected by a Twitter streaming API. Our study period ran from August 18 to September 12, 2017, covering the time before (August 18–24), during (August
Discussion
In this paper, we have presented an urban analytical and geo-visual system that automatically collects geotagged tweets and performs urban event detection and visualization in real time. Through the application in the simulated streaming data from a large-scale hurricane, the system has been demonstrated to provide useful and timely information on emergencies and crises during disasters. The system is specifically designed to address the smart rapidity aspect of resilient cities by enabling
Conclusion
Building resilient cities requires smart solutions, and achieving smart rapidity is one of the most important approaches to enhancing urban resilience when promoting smart cities. We develop a real-time urban analytical and geo-visual system for social media streaming data to track small-scale undefined urban extreme events and provide early emergency alerts. The system has demonstrated the effectiveness and rapidity in processing large volumes of data with low latency. The system has the
Declaration of Competing Interest
The authors report no declarations of interest.
Acknowledgements
This material is based upon work supported by the National Science Foundation under Grant No. 1760645, Grant No. 2028012, Grant No. 1951816, and University of Florida DCP Research Seed Grant Initiative. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation or University of Florida.
References (74)
- et al.
Public behavior response analysis in disaster events utilizing visual analytics of microblog data
Computers and Graphics (Pergamon)
(2014) - et al.
Designing, planning, and managing resilient cities: A conceptual framework
Cities
(2013) - et al.
Generation of topic evolution graphs from short text streams
Neurocomputing
(2020) - et al.
Social media big data analytics: A survey
Computers in Human Behavior
(2019) - et al.
Real-time event detection from the Twitter data stream using the TwitterNews+ Framework
Information Processing and Management
(2019) - et al.
A cloud-enabled automatic disaster analysis system of multi-sourced data streams: An example synthesizing social media, remote sensing and Wikipedia data
Computers, Environment and Urban Systems
(2017) - et al.
The resilience to emergencies and disasters index: Applying big data to benchmark and validate neighborhood resilience capacity
Sustainable Cities and Society
(2018) Climate change and urban resilience
Current Opinion in Environmental Sustainability
(2011)- et al.
Gauging urban resilience from social media
International Journal of Disaster Risk Reduction
(2018) - et al.
A methodology for real-time data sustainability in smart city: Towards inferencing and analytics for big-data
Sustainable Cities and Society
(2018)