Florida Division of Elections Scanners
note: This is a side project and not as rigorous as it should be, yet.
These Ruby scripts scrape the Florida Division of Elections database for info on campaign contributions. Shoutout to the Palm Beach Post for starting this project.
For a good time, point the scraper to collect contributions between the year 9000 and 10,000.
Political committees and campaign expenses are a work in progress.
Getting Started
Assuming you have Ruby, install Dependencies
gem install bundler
bundle install
Option 1: Straight to CSV
Run ruby scrapers/contrib_csv_generator.rb to gather contribution information.
The data will be organized for easy import into a Neo4j database, but you can naturally modify the output as you wish.
Option 2: Into a Neo4j database
After importing the data to CSV, import it using the Neo4j import tool (see below), then use $NEO4J_HOME/bin/neo4j console to start the database.
$NEO4J_HOME/bin/neo4j-admin import --ignore-duplicate-nodes=true --ignore-extra-columns=true --id-type=STRING --delimiter="\t" --ignore-missing-nodes --nodes:Candidate headers/candidateNodeHeader.csv,output/candidates.* --nodes:Committee headers/committeeNodeHeader.csv,output/committees.* --nodes:Contribution headers/contributionCo.csv,output/committees.* --nodes:Contribution headers/contributionCa.csv,output/candidates.* --nodes:Contributor headers/contributorCa.csv,output/candidates.* --nodes:Contributor headers/contributorCo.csv,output/committees.* --relationships:CONTRIBUTED headers/contributedCa.csv,output/candidates.* --relationships:CONTRIBUTED headers/contributedCo.csv,output/committees.* --relationships:SUPPORTS headers/supportsCo.csv,output/committees.* --relationships:SUPPORTS headers/supportsCa.csv,output/candidates.*
Nodes
- Contributor
- Candidate
- Contribution
- Committee
Relationships
- SUPPORTS
- CONTRIBUTED
Sample queries
Who/what contributed the most (in dollars), and to who?
The query below assumes a Party node and a BELONGS_TO relationship.
First, create the relationship:
MATCH (candidate:Candidate)
MERGE (candidate)-[:BELONGS_TO]->(p:Party)
SET p.name = c.party
Then run the query
MATCH (contributor:Contributor)-->(c:Contribution)-[:SUPPORTS]->(candidate:Candidate)-[:BELONGS_TO]->(party:Party)
RETURN contributor.name, count(c) AS count, sum(c.amount) AS amount, collect(distinct candidate.name), collect(distinct party.name) AS Party, collect(distinct c.year)
ORDER BY AMOUNT DESC
LIMIT 10
Who contributed the most (in number of contributions), and to who?
MATCH (contributor:Contributor)-->(c:Contribution)-[:SUPPORTS]->(candidate:Candidate)-[:BELONGS_TO]->(party:Party)
RETURN contributor.name, count(c) AS count, sum(c.amount) AS amount, collect(distinct candidate.name), collect(distinct party.name) AS Party, collect(distinct c.year)
ORDER BY COUNT DESC
LIMIT 10
What city made the most contributions, and to who?
MATCH (contributor:Contributor)-->(c:Contribution)-[:SUPPORTS]->(candidate:Candidate)-[:BELONGS_TO]->(party:Party)
WITH c.city AS city, count(distinct contributor) AS contributors, count(c) AS contributions, collect(distinct party.name) AS party
RETURN city, contributors, contributions, party
ORDER BY contributions DESC
LIMIT 10
How do the big names play the field?
Try searching for names that contain GEO GROUP or variations of AT&T
MATCH (geo:Contributor)-->(c:Contribution)-[:SUPPORTS]->(candidate:Candidate)
WHERE geo.name CONTAINS "AT&T" OR geo.name contains "AT &T"
RETURN collect(distinct geo.name), candidate.party AS party, count(distinct candidate.name) AS candidates, sum(c.amount) AS amount