WorldBoard

• 

  a poster for WorldBoard

• 

  a screenshot of the webapp

• 

  an analysis of withdrawal data showing an unusual spike

Inspiration

• we wanted some sort of dashboard for world events
• we came across the GDELT database -- live events updated every 15 minutes

Analysis

• we looked at CapitalOne API finance data to find volatile banking activity which we expected might reflect the state of the world
• we extracted data from the GDELT database and analyzed with Excel and R to find relevant world news trends and other meaningful features

Product

• a user-friendly web-based dashboard for analyzing the last fifteen minutes of GDELT data
• a good location for finding news articles that are important and legitimate in content but grossly overlooked by mainstream media and feeds

Future

• the easy ability to add more feeds to the program
• more feature extraction and clustering and unsupervising learning to extract meaningful trends
• automatic or live correlations between news and finance, twitter, etc. feeds

Built With

• atom
• canva
• css
• docker
• excel
• excel-macro
• flask
• gdelt
• git
• golden-layout
• highcharts
• javascript
• json
• node.js
• npm
• numpy
• pandas
• python
• r
• rstudio
• vim
• webpack