Twitter Crawler is a Neo4j experiment. It is as simple as possible with no intention
to model any serious application architecture or something.
It recursively fetches Twitter friends (the users a user follows) up to to a limited depth and width
and enriches this data with geo location from Yahoo Places.

I did not bother making it configurable without code modification, since it is not a library
or a real application.

Configure
---------

- Put the path to your Java JDK into the POM. 1.7 required.
- Open Config.java and modify the configuration to your needs. Your Twitter ID can be
  fetched via the Twitter API developer console or with this call:
  https://api.twitter.com/1/users/lookup.xml?screen_name=<your twitter name here>

Run
---

- Run the test CrawlerTest.testFriendCrawl
  This will collect the Twitter friend graph into the Neo4J DB and cache all Twitter data in the files
  which you have configured in Config.java.
  Twitter limits the API calls to 150 per hour. The crawler stops when Twitter refuses the requests.
  Restart it as often as you like. The Neo4j DB can always be regenerated on demand with the contents
  of the cache files.
- Run the test CrawlerTest.testDataFill
  This will enrich the Twitter graph with user information and get geo data from Yahoo.

Visualize
---------

- Download Gephi with the Neo4J Plugin. Enjoy the Geo Layouter!