Wed, 03 Oct 07

Finding similar sites by mining the data

So this is very much proof of concept and definitely not very useful in its current form…

You need to execute this javascript in the context of a webpage. In firebug (for example), paste it into the console and run.

// Execute this within an existing webpage, using, for example, firebug console
function includeDelicious() {
  function loadScript(scriptUrl) {
    var scriptTag = document.createElement('script');
    scriptTag.setAttribute('type', 'text/javascript');
    scriptTag.src = scriptUrl;
if (typeof(getRelatedPosts) != 'function') { includeDelicious(); }

The javascript above loads this script that in turn loads md5.js, jquery.js, does some magic with the json feeds before finally writing some ‘similar’ sites to the bottom of the page you’re currently looking. At the moment it finds the three most recent sites you’ve bookmarked for each of the tags that people have used when bookmarking the site you’re looking at. It writes all this to the bottom of the current page, as this screenshot shows.

Lots of things to do before this could possibly be deemed useful but I think it’s got some potential.