Near-duplicate detection for news stories. Pulls a day's worth of stories from a MongoDB instance, uses the simhash algorithm to cluster the stories, and outputs the clustered data to a file.
Requires docker and docker-compose. Once both of those are installed:
docker-compose up -d
Where the -d flag indicates for the process to run in the background.