gistfile1.rb

require 'logger'

class String
  LOGGER = Logger.new($stderr)
  def -@; LOGGER.info self; end
end

# Extracted earlier, hardcoded for speed
ENTITIES = %w(I-ORG O I-MISC I-PER I-LOC B-LOC B-MISC MO B-ORG)

- 'started parsing training data'

TRAINING_DATA = CONLL03.parse_file(STDIN)

- 'finished parsing training data'

model = Perceptron.new ENTITIES, CONLL03.labels