javascriptでなんとなく単語を抽出する

gistfile1.js

(function() {
  var tags = [];
  var _tags = {};
  var words = document.documentElement.innerText.match(/[\u3400-\u4DBF\u4E00-\u9FFF\uF900-\uFAFF]|[\uD840-\uD87F][\uDC00-\uDFFF]+|[\u12449-\u12531\u12540;]+|[a-zA-Z0-9!]+|[ａ-ｚＡ-Ｚ>０-９]+/g);
  words.forEach(function(word) {
  if (word.length <= 3 || _tags[word]) { return ; }
  _tags[ word ] = 0;
  tags.push( word );  });
  delete _tags;
  return tags;
})();