- We want to make the stories of refugees more comprehensible
What it does
- It empowers journalists and people involved with the refugee crisis in storytelling.
- We display terror data on a world map for the user to explore.
- We provide additional information like statistics for the region, fitting New York Times articles and casualties.
How we built it
- We downloaded public datasets that contained a list terror attacks, riots, explosions and violence in general. We also downloaded a dataset that describes refugee movement throughout the last few years.
- We used Python and Jupyter Notebooks to pre-process the data and bring it into a normalized format. Every 'event' (riot, explosion, ...) now looks like so:
"location": ["Greece","Attica","Athens","Central Athens","Athens-Central Athens"],
"description": "On 12 May, just after midnight, groups of hooded youths [allegedly anarchists] lobbed Molotov cocktails at a police detachment in Exarchia near the Ministry of Culture and later at a platoon of riot policemen guarding the offices of ruling party SYRIZA in central Athens. Three cars and a motorcycle were damaged. Police responded with tear gas. Seven people were detained but released later. [size=no report]"
Challenges we ran into
- data normalization
- finding a free to use API for additional information like social media or newspaper
Accomplishments that we're proud of
- mastering d3.js
- realizing our own interpretation of the challenge
What we learned
- Data normalization is more difficult and time-consuming than it looks (especially when it comes to country names, country codes and geo data)
- Visualizations with d3.js are hard. Really hard.
- Surprisingly, the New York Times have a free API to search all past articles.
What's next for Refugee Movement: Cause and Effect
- Present our tool to refugee workers and journalists and find out what else they wish for.
- Collaborate with them to create an interactive experience with a specific story.
- Ideally publish it.
- Sleeping! Good night!