Ankusa::NaiveBayesClassifier

# File lib/ankusa/naive_bayes.rb, line 13
def classifications(text, classnames=nil)
  result = log_likelihoods text, classnames
  result.keys.each { |k|
    result[k] = (result[k] == INFTY) ? 0 : Math.exp(result[k])
  }

  # normalize to get probs
  sum = result.values.inject { |x,y| x+y }
  result.keys.each { |k| result[k] = result[k] / sum }
  result
end

# File lib/ankusa/naive_bayes.rb, line 7
def classify(text, classes=nil)
  # return the most probable class
  log_likelihoods(text, classes).sort_by { |c| -c[1] }.first.first
end

# File lib/ankusa/naive_bayes.rb, line 26
def log_likelihoods(text, classnames=nil)
  classnames ||= @classnames
  result = Hash.new 0

  TextHash.new(text).each { |word, count|
    probs = get_word_probs(word, classnames)
    classnames.each { |k| 
      # log likelihood should be infinity if we've never seen the klass
      result[k] += probs[k] > 0 ? (Math.log(probs[k]) * count) : INFTY
    }
  }

  # add the prior
  doc_counts = doc_count_totals.select { |k,v| classnames.include? k }.map { |k,v| v }
  doc_count_total = (doc_counts.inject { |x,y| x+y } + classnames.length).to_f
  classnames.each { |k| 
    result[k] += Math.log((@storage.get_doc_count(k) + 1).to_f / doc_count_total) 
  }

  result
end