Take a look at the the demo before reading : 
// http://stackoverflow.com/a/3561711/1636522
RegExp.escape = function(s) {
  return s.replace(/[-\/\\^$*+?.()|[\]{}]/g, '\\$&');
};
// vars
var span = getEl('span'),
  input = getEl('input'),
  li = getEls('li'),
  tid;
  
// onkeyup
addEventSimple(input, 'keyup', function(e) {
  // cancels previous query
  tid && clearTimeout(tid);
  // waits 250ms then filters
  tid = setTimeout(function() {
    tid = null;
    span.textContent = +span.textContent + 1;
    filter(e.target.value);
  }, 250);
});
// filtering
function filter(input) {
  var i = 0,
    l = li.length,
    re = input && toRegex(input),
    el;
  for (; i < l; i++) {
    el = li[i]; // list item
    if (!re || re.test(el.textContent)) {
      el.style.display = 'list-item';
    } else {
      el.style.display = 'none';
    }
  }
}
// input > regex
function toRegex(input) {
  input = RegExp.escape(input);
  input = input.match(/[^,\s]+(\s+[^,\s]+)*/g) || [];
  input = input.join('|');
  return new RegExp(input, 'i');
}
// http://www.quirksmode.org/js/eventSimple.html
function addEventSimple(obj, evt, fn) {
  if (obj.addEventListener) obj.addEventListener(evt, fn, false);
  else if (obj.attachEvent) obj.attachEvent('on' + evt, fn);
}
// helpers
function getEl(tag) {
  return getEls(tag)[0];
}
function getEls(tag) {
  return document.getElementsByTagName(tag);
}
<input type="text" placeholder="Example : "nn, oo, ca"." />
<div style="padding:.5em .5em 0">Filtered <span>0</span> times.</div>
<ul>
  <li>Guybrush Threepwood</li>
  <li>Elaine Marley</li>
  <li>LeChuck</li>
  <li>Stan</li>
  <li>Voodoo Lady</li>
  <li>Herman Toothrot</li>
  <li>Meathook</li>
  <li>Carla</li>
  <li>Otis</li>
  <li>Rapp Scallion</li>
  <li>Rum Rogers Sr.</li>
  <li>Men of Low Moral Fiber</li>
  <li>Murray</li>
  <li>Cannibals</li>
</ul>
 
 
Here I'll only expose the toRegex function. Say that we have entered the following value : "el, le, az".
var regex = toRegexp('el, le, az'); // regex = /el|le|az/i
regex.test('Elaine'); // true  -> show
regex.test('Marley'); // true  -> show
regex.test('Stan');   // false -> hide
The resulting regular expression (/el|le|az/i) means : search for "el" or "le" or "az", and ignore the case (allows "EL", "Le" or "aZ" as well). Now, let's read this function line by line :
input = RegExp.escape(input); // http://stackoverflow.com/q/3561493/1636522
input = input.match(/[^,\s]+(\s+[^,\s]+)*/g) || []; // ["el", "le", "az"]
input = input.join('|'); // "el|le|az"
return new RegExp(input, 'i'); // /el|le|az/i
Let's go further about /[^,\s]+(\s+[^,\s]+)*/g :
[^,\s]+         any char except comma and whitespace, one or more times
(\s+[^,\s]+)*   one or more whitespaces + same as above, zero or more times
g               grab all occurrences
Usage example with a stupid input :
'a,aa,aa a, b , bb , bb b , , '.match(/[^,\s]+(\s+[^,\s]+)*/g);
// ["a", "aa", "aa a", "b", "bb", "bb b"]
That's it ! Hope that was clear enough :-)
Further reading : http://www.javascriptkit.com/javatutors/redev.shtml.