(Geo-)Social Media and Mapping: Which Spatial Footprints Do We Leave via Online Services?

At Mapillary, we are all in for research and innovation. In this guest post, Levente Juhász describes his academic research about Volunteered Geographic Information and asks for some help from our community.

At Mapillary, we are all in for research and innovation. In this guest post, Levente Juhász describes his academic research about Volunteered Geographic Information and asks for some help from our community.

Note: this is Levente’s own PhD research, not a project by Mapillary.

Some of you might remember my previous visualizations and blog posts, such as the one about Mapillary’s rhythm, or when I talked about how I usually map hiking trails with Mapillary. Lately, I have been focusing on another aspect of my life—earning a PhD at the University of Florida. I’m happy to report that I’ve come close to the end, but there’s still one last piece of research I would like to do. A final research project that I can’t complete without your help.

I am looking at how the same individual uses different online platforms simultaneously, and how this is reflected in the spatial footprints of those contributions. So far, almost no researchers have tried to understand this activity of users, so there is plenty to do. I’m hoping that looking at user-generated data from multiple sources will provide us with some insights of human mobility, and can help better understand how we interact with space while on social media.

Mapillary mapping gear

A Mapillary photo-shooting setup, posted on Instagram

Sounds exciting? Read more below and I hope that you get as excited about it as I am. If you want to cut directly to the chase, visit my research page at https://research.jlevente.com and learn about the project from there.

Main research topic—Volunteered Geographic Information

Entering the era of Web 2.0 brought lots of technological developments. These days, we generate lots of data, from messaging our friends to uploading a whole album of holiday photos to an online repository. While we do this, we interact with the space around us on so many levels. We mention places we meet at, we use navigation with real-time traffic reports, and so on. There is a smartphone in almost every pocket capable of precisely locating yourself wherever you are on Earth.

In GIScience, data with a geographic component that has been generated by regular people is called Volunteered Geographic Information (VGI). Think of a geotagged Instagram photo, a tweet, a Strava workout, an OpenStreetMap (OSM) edit, or your favorite Mapillary sequence. Over the last decade, VGI has become a hot topic since it provides easy access to massive datasets and helps researchers answer a number of questions effectively.

To name a few, scientists have looked at human mobility [1] through the lens of Twitter, identified tourist hot spots from Panoramio [2], and determined where bicyclists ride for fitness using Strava data [3]. Since the term was first coined in 2007, Google Scholar shows more than 20,000 research outputs related to VGI.

A new way of looking at VGI

This enormous research effort in studying VGI resulted in a good understanding of how people tweet, how they edit OSM, and so on. However, most VGI studies have been concerned with a single platform, and therefore, our knowledge of how the same individual uses these services simultaneously is limited at best.

And the thing is: people do use many online services during their everyday lives. I know for a fact that I take Mapillary photos, edit OSM, record my bike rides, and post social media photos. I believe many people do the same and while doing so, these activities result in different spatial footprints. This is what I’m mainly interested in.

My research aims to advance our understanding of this topic I call cross-platform user behavior analysis. I believe that a better understanding of how people interact with space while on different online services can bring us a number of benefits, for example, by building more accurate mobility models, or by creating more effective detection systems of local events.

Example of how a person uses Mapillary to edit OSM

Current work and how you can help

To get to the point of understanding, I need to look at more than just a few people’s online activities. In an earlier study this year, I already explored tools and methods to compare the overlap of social media activities between different platforms, and to mathematically quantify (dis)similarity between point patterns [4]. This pilot study will be presented at AGILE 2018 in Lund. Make sure to come say hi if you’re around!

Social media activities within a city Example of a user’s social media activities within a city [4]

While useful, the pilot study was limited to the contribution of only a few users and just two popular social media services. The next step is taking this research further by coming up with a comprehensive evaluation of how this phenomenon takes place.

How am I going to achieve this, you might ask. Well, I’m relying on the power of community, and on you. I’m looking for awesome people who would voluntarily help me by sharing their online profiles and thus contribute to the successful completion of my research.

I built a website that allows you to authenticate with 10 social media and mapping platforms and authorize my tools to extract some location information from your profiles. Your participation is entirely voluntary, but it would mean a lot to the success of this project.

The services you can use are Twitter, Instagram, Facebook, Strava, iNaturalist, Flickr, Meetup, and Foursquare. Since open data and mapping are special to me, you can also authenticate with Mapillary and OpenStreetMap. Visit my site at https://research.jlevente.com if you are interested.

Consider contributing to my research and help advance our understanding of cross-platform user behavior. Even if you are not big on social media, you could still help by sharing my project with others who might be interested. It would be a big help.

Connecting different social accounts Connecting different social accounts on my research page

A special note on privacy

I realize the sensitiveness of location data, therefore I have taken measures to protect your privacy. The whole website is built on top of Django. The authentication and authorization process uses the standard OAuth flow, meaning that your passwords are NOT shared with me but only with the provider. The site is secured with SSL. Your tokens and data are also stored in different databases.

I will use your data solely for my academic research. It will not be disclosed, shared with or sold to any third parties. My approach is to be open and transparent about what I do. You can read more about the research and technical details (including a complete list of information I collect and ways to opt out, etc.) at https://research.jlevente.com. With any questions or concerns, you can email me at levente.juhasz@ufl.edu and I’m more than happy to discuss.

If you got to this point you already put a lot of effort in learning more about my research. I am grateful for that. I hope that you find it as exciting as I do and you will consider helping out.




[1] Hawelka, B., Sitko, I., Beinat, E., Sobolevsky, S., Kazakopoulos, P., & Ratti, C. (2014). Geo-located Twitter as proxy for global mobility patterns. Cartography and Geographic Information Science, 41(3), 260-271.

[2] García-Palomares, J. C., Gutiérrez, J., & Mínguez, C. (2015). Identification of tourist hot spots based on social networks: A comparative analysis of European metropolises using photo-sharing services and GIS. Applied Geography, 63, 408-417.

[3] Griffin, G. P., & Jiao, J. (2015). Where does bicycling for health happen? Analysing volunteered geographic information through place and plexus. Journal of Transport & Health, 2(2), 238-247.

[4] Juhász, L. and Hochmair, H. H. (2018). Cross-checking user activities in multiple geo-social media networks. 21st AGILE Conference on Geo-information Science. Lund, Sweden

Continue the conversation