In 2018 alone, almost 120,000 conflicts occurred with millions of people suffering. With our world becoming more global, borders are dissolving. Static maps only depict a fraction of the information and fail to visualize movement and dependencies of events. We need to empower decision-makers to understand conflict development in order to enhance coordination of humanitarian help.
What it does
Visualizing how conflicts and actors move, providing trend lines and analytical tools to extract knowledge for action. Shows: 1) Monthly shift of incident location 2) chronological order 3) links events to actors 4) highlights cross-border operations of actors 5) terrorist group's areas of influence 6) correlated actors 7) categorization of conflict type over time 8) Per actor: type of conflict
How we built it
Backend using data processing with Python, Frontend visualization using D3.js and leaflet.js
Challenges we ran into
Many data sets lack detailed data (only year, country). Fancy visualizations more complex than initially throught. Maps quickly become crowded with markers (need for grouping&filtering).
Accomplishments that we are proud of
Working visualization. Uncovering new correlations in the new map that could not be extracted with previous visualizations. Enhanced usability through the implementation of map navigation and automated pan on the sidebar.
What we learned
Visualization is hard. Difficult to include many information units without sacrificing clarity. Data mining did not provide useful insights, more time for model configuration needed.
What's next for Unbounded
Connect further data sources for advanced information & Move from reporting to predicting