Network data visualization

The Master’s degree in Intelligence Technologies and Digital Design at the Pontifical Catholic University of São Paulo (PUC-SP) allowed me to develop this project which is very dear to me. Basically, I built some scripts to download data from social networks and build relationship-based maps (or visualizations, or topologies…) to detect influencers, messages that propagate better online and how digital communities are formed about a theme.

I used this methodology to analyze the interactions between voters during the Brazilian presidential election in 2018 for seven straight months, with weekly reports. When elections are taken to social network platforms, there is a high level of entropy in the exchange of the messages, so the data visualization comes in handy to see the information fluxes and also to detect anomalies.

It has recently been confirmed that bots were key to elect the actual Brazilian president. Elsewhere in the world, message targetting and bots are said to play a major role in the American elections, for instance. This is why I find creating these maps of utmost importance..

Up-to-date maps can be seen at the project’s Facebook page: https://www.facebook.com/afinimapa

Here are some examples:

1. The finished product accompanying a 15-page report
2. A raw image showing influence zones (delimited by semantic space)
3. Bot detection (bots in red)