[Paper Review] Micromobility in Smart Cities: A Closer Look at Shared Dockless E-Scooters via Big Social Data

Yunhe Feng, Dong Zhong+5 more|arXiv (Cornell University)|Oct 28, 2020

Human Mobility and Location-Based Analysis11 references25 citations

TL;DR

This study leverages 5.8 million geotagged, scooter-related tweets and 144,197 images from October 2018 to March 2020 to conduct a large-scale, multi-dimensional analysis of shared dockless e-scooter usage in smart cities. Using big social data analytics, it profiles spatial-temporal usage patterns, identifies key stakeholders (riders, gig workers, companies), examines injury and parking behaviors, and performs sentiment analysis, revealing a 65.14% male and 34.86% female rider gender gap, widespread improper parking (62.61%), and predominantly positive public sentiment despite safety concerns.

ABSTRACT

The micromobility is shaping first- and last-mile travels in urban areas. Recently, shared dockless electric scooters (e-scooters) have emerged as a daily alternative to driving for short-distance commuters in large cities due to the affordability, easy accessibility via an app, and zero emissions. Meanwhile, e-scooters come with challenges in city management, such as traffic rules, public safety, parking regulations, and liability issues. In this paper, we collected and investigated 5.8 million scooter-tagged tweets and 144,197 images, generated by 2.7 million users from October 2018 to March 2020, to take a closer look at shared e-scooters via crowdsourcing data analytics. We profiled e-scooter usages from spatial-temporal perspectives, explored different business roles (i.e., riders, gig workers, and ridesharing companies), examined operation patterns (e.g., injury types, and parking behaviors), and conducted sentiment analysis. To our best knowledge, this paper is the first large-scale systematic study on shared e-scooters using big social data.

Motivation & Objective

• To understand the spatial and temporal dynamics of shared dockless e-scooter usage across major cities using real-time social media data.
• To identify and profile key stakeholders, including riders, gig workers, and ridesharing companies, through social media engagement.
• To examine operational challenges such as injury types, parking behaviors, and regulatory concerns via crowdsourced data.
• To conduct a comprehensive sentiment analysis of public perception toward e-scooter sharing services using textual, visual, and emoji-based signals.
• To provide the first large-scale, systematic study of shared e-scooters using big social data, offering actionable insights for urban planners and policymakers.

Proposed method

• Collected 5.8 million English-language tweets mentioning 'scooter' or the scooter emoji via Twitter Streaming API from October 2018 to March 2020.
• Performed multi-granularity geospatial and temporal analysis on tweet distributions at city and hourly levels to identify usage trends.
• Applied topic modeling (LDA) and hashtag analysis to extract and categorize 12 prominent discussion topics, including regulations, injuries, and market shares.
• Used computer vision and image recognition to detect e-scooter brand logos in 144,197 user-uploaded images to estimate market share and assess parking and vandalism behaviors.
• Conducted sentiment analysis using TextBlob polarity scores, facial emojis, and emoticons to classify public sentiment into positive, neutral, and negative categories.
• Combined app screenshot analysis with text and GPS data to estimate median trip duration and cost, and to classify injury types and parking locations.

Experimental results

Research questions

• RQ1 How do e-scooter usage patterns vary across cities and over time in terms of tweet volume and geographic distribution?
• RQ2 What are the dominant topics of public discussion on social media regarding shared e-scooters, and how do they relate to stakeholders, operations, and emotions?
• RQ3 What are the gender demographics of e-scooter riders based on social media self-reports, and what are the implications for equity and access?
• RQ4 What are the most common injury types and parking behaviors reported by users, and how do they vary by location?
• RQ5 What is the overall public sentiment toward shared e-scooter services, and how do linguistic cues (words, emojis, emoticons) reflect positive or negative perceptions?

Key findings

• The United States, New Zealand, and Australia showed a decreasing trend in e-scooter-related tweets, while Canada and India exhibited increasing usage, with the UK remaining stable.
• A gender gap in e-scooter ridership was confirmed, with 65.14% of self-identified riders as male and 34.86% as female based on social media data.
• 62.61% of e-scooters were parked improperly, with 34.78% blocking sidewalks, 17.39% in other wrong areas, 5.65% vandalized, and 4.78% parked indoors.
• The most common injury types were to the legs and feet (49.67%), followed by the trunk and hands (27.45%), and head (22.88%), with 5.88% of injuries involving the face.
• Public sentiment was predominantly positive: 3.48 times more positive than negative polarized words were used, and the face with tears of joy emoji was the most frequently used facial emoji.
• The median trip cost was estimated at $3.00, and the median trip duration at 12.5 minutes, based on analysis of e-scooter app screenshot data.