Towards resilient and smart cities: A real-time urban analytical and geo-visual system for social media streaming data

Highlights

• A real-time system for analyzing and visualizing streaming data from social media.

• Online topic modeling and sentiment analysis with Apache Spark distributed system.

• Interactive real-time geo-visualization using GeoEvent Server and Online GIS.

• Intelligent and Integrated geo-visual data analytics for smart and resilient cities.

• Rapid events detection and tracking during disasters and public health crises.

Abstract

Cities worldwide are vulnerable to unpredictable extreme events such as disasters and public health crises. Urban big data and data-driven technologies have played an increasingly important role in building smart and resilient cities that can respond rapidly to these perturbations. However, many existing approaches had limited capabilities for processing big data, which has led to time-consuming and costly decision-making. Thus, we develop a real-time data-driven analytical and geo-visual system to enable smart and rapid responses to urban extreme events. The system is built on ArcGIS’s GeoEvent Server and Apache Spark and processes streaming data from social media with high speed, massive volume, and multiple modalities. The system employs online topic modeling and domain-adaptive sentiment analysis to track small-scale, undefined events, visualizes their spatial and semantic dynamics, and provides early alerts for crises and emergencies via an interactive online GIS platform. The proposed system has been applied during a large-scale hurricane and demonstrated effectiveness and agility in tracking and reporting emerging small-scale crises. The developed system can be applied in various urban scenarios to enable timely situation awareness and rapid response. This research contributes to the smart city safety and building rapidity of resilient cities.

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.