Abstract

Crowdsourcing platforms and social media produce distinctive geographies of informational content. The production process is enabled and influenced by a variety of socio-economic and demographic factors, shaping the place representation, i.e., the amount and type of information available in an area. In this study, we explore and explain the geographies of Twitter and Wikipedia in Greater London, highlighting the relationships between the crowdsourced data and the local geo-demographic characteristics of the areas where they are located. Through a set of robust regression models on a sample of 1.6M tweets and about 22,000 Wikipedia articles, we identify level of education, presence of people aged 30–44, and property prices as the most important explanatory factors for place representation at the urban scale. To some extent, this confirms the received knowledge of such data being created primarily by relatively wealthy, young, and educated users. However, about half of the variability is left unexplained, suggesting that a broader inclusion of potential factors is necessary.

Highlights

  • Over the past decade, the diffusion of crowdsourcing platforms and GPS-enabled smartphones has enabled the large-scale production of spatial information

  • Ballatore and De Sabbata nomenon has been variously characterised as spatial crowdsourcing, volunteered geographic information (VGI), spatial social media, and user-generated content (UGC) (Sui et al, 2012; See et al, 2016)

  • We investigated the information geographies of two well-known crowdsourced data sources, Twitter and Wikipedia, observing their place representation in Greater London

Read more

Summary

Introduction

The diffusion of crowdsourcing platforms and GPS-enabled smartphones has enabled the large-scale production of spatial information. Studies of information geographies have so far focused on large spatial units, such as countries (Graham et al, 2015a), with few works focusing on the urban or regional scale. The latter is of particular importance, as VGI tends to be produced in urban areas (Hecht and Stephens, 2014). When using VGI, it is necessary to consider the socio-spatio-temporal processes that supported its generation, thinking about what data is missing, and about what is visibly present

Objectives
Findings
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call