Similicio.us - Crowdsourcing Similar Sites
Similicio.us is a very simple search engine, dedicated to a single task… Find similar sites. Basically, you provide similicio.us with a domain name, and it will return other sites similar in function or content. Now, at first glance this service is nothing new. After all, Google has been including their “similar pages” link in their search results for years. But - after I had a chance to use similicio.us - I found it returns some promising results…. sometimes unexpected, but promising none-the-less.
Before I talk about the functionality of the similicio.us, it’s important to mechanics behind the site. Similicio.us - as the name cleverly implies - uses Yahoo!’s del.icio.us social bookmarking service as it’s primary datasource. The site queries del.icio.us with the URL you’re looking for and returns entries with the same tags. This is a rather creative way to use crowdsourcing as a basis for a search engine. After all, this is what the concept of “tagging” was originally designed for.
To the Test!
Ok, so now it’s time to put it through the paces and see how well it is at finding similar results. The first couple of sites were simple… Cnn.com returned Reuters, BBC, Wired and other news sites - While Etsy.com returned Instructables.com, elsewares.com, threadless.com, and crafters.org… so far so good.
Now, I have found a few flaws in using the crowdsourcing approach. As an example, I did a search for “Twitter.com” that returned some interesting results. For those who are not familiar Twitter is an application that most people would classify as a “Micro Blogging” or “Lifecasting” type of application. Essentially, it allows you to post your thoughts, activities, questions… really anything you want in 160 characters or less. So, you would think Similicio.us would return results like Plurk.com, Wordpress.com, or Blipper.com which are similar applications. However, what I found is that it returned sites like TechCrunch.com, ReadWriteWeb.com, Google.com, and other odd selections. These sites really don’t have anything in common with Twitter - Besides some being actual blogs. But what I realized is that these sites have a lot of content about Twitter (and more specifically it’s downtime problems).
So, there is an inherent flaw associated with tagging, However the majority of the searches did bring back promising results. I did notice that Addoursearch.com returned zero results, but I won’t hold that against them :)
