http://www.catalysoft.com/articles/StrikeAMatch.html

word_similarity.rb

words = {
  good: "france",
  bad: "french"
}

pairs = {
  good: [],
  bad: []
}

words.each do |kind, word|
  buffer = ""

  word.each_char do |char|
    pairs[kind] << buffer + char
    buffer = char
  end
end

remainder = (pairs[:good] - pairs[:bad])
result = (remainder.length * 2.0) / (pairs[:good].length + pairs[:bad].length)
#=> 0.5

puts result