Skip to content

kmcurry/story-where

Repository files navigation

story-where

Given a corpus of news stories:

  • How precisely can we identify geographic features in news stories?
    • place names
    • longitudes, latitudes
    • rural, suburban, urban
  • Can we put identified geographic features into context?
  • Where are stories located on maps?
  • How are stories distributed by geographic features?
  • Are there reasonable correlations between the geographies of news stories and other boundaries?
    • Census Tracts
    • Zip Codes
    • School Zones
    • Voting Districts

About

Story Where uses a corpus of news articles to extract locations and map them. Story Where is being written with 2 years of articles exported from The Virginian-Pilot's content management system. Roughly 49K articles were processed into our pilot database.

Setup

This web application is built on Google Cloud with Postgres and Python 3.

We referenced the "Writing a Basic Web Service of App Engine" guide to get started.

See web-app/requirements.txt for dependencies.

Mac users on Mojave may encounter an error with psycopg2, which is used by SQLAlchemy. The error can be resolved by running the following: sudo installer -pkg /Library/Developer/CommandLineTools/Packages/macOS_SDK_headers_for_macOS_10.14.pkg -target / reference

Keys

The following keys are required:

  • server-ca.pem
  • client-key.pem
  • client-cert.pem

Environment Variables

The following env vars are required:

  • PG_USE_SSL
  • PG_CONN_STR
  • PG_CLIENT_KEY
  • PG_CLIENT_CERT
  • PG_SERVER_CA
  • GOOGLE_APPLICATION_CREDENTIALS
  • MAPBOX_KEY

About

Given a corpus of news stories, where are they on maps and how precisely can we map their place names?

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •