This code requires a number of libraries to run:

Lucene (http://lucene.apache.org/core/)
JDBC for PostgreSQL
Stanford NER Library (http://nlp.stanford.edu/software/CRF-NER.shtml)

The following SQL commands should be run before this code is run for the first time:

drop table user_vertices;
drop table tweet_vertices;
drop table edges;
drop table edges_tweetids;

create table user_vertices (name varchar(32) primary key, score float, original_score float);
create table tweet_vertices (tweetid int primary key, name varchar(32), date timestamp, tweet varchar(255), score float, original_score float);
create table edges (name varchar(32), tweetid int, type int);

create index idx_edge_name on edges using btree(name);
create index idx_edge_tweet on edges using btree(tweetid);
create index idx_edge_type on edges using btree(type);
create index idx_tweet_author on tweet_vertices using btree(name);