Home

Tips for using Stumble Upon

My name is Robin Gower and I'm a Stumble Upon addict. I'd like to encourage you to take-up this pastime and to share with you a few of the tips that I've learned.

There is a growing clamour around the idea of a second coming of the internet, Web 2.0 to give the meme it's title. One aspect is the social interpretation (or indexing) of the web. There are a range of site that draw upon the wisdom of crowds by aggregating users' votes to recommend content.

Google's page rank mechanism treats links as votes. Social voting sites are just as simple, but by counting conscious choices rather than references (links) these sites have a more precise focus on the data that identifies good content. Unlike page rank, social votes can be for or against (Thumbs up vs Thumbs down). This makes these mechanisms more informative still by weeding-out bad pages or junk data.

Digg is sensationalist. The site is clogged by exaggerated polemics so that it reads like the front page of a tabloid. Although I seldom visit the site, I still use the Digg iGoogle widget but it rarely attracts my attention.

Stumble Upon offers all of the benefits of Digg and adds a further feature. In SU content recommendations also draw on correlations between users, rather than just counting total number or rate of votes. SU weights votes according to the preferences expressed automatically by the user. Again a simple distinction leads to a dramatic improvement in the quality of information.

Priming Stumble Upon
I started doing this after I'd been using the toolbar for a few weeks. Basically you can give SU a more rapid insight into your interests by telling it first (rather than waiting for it to discover these organically). You can prime SU by:

  • Of course you're asked to choose from a finite set of categories when you first register. It might be tedious but completing this task fully will save you time in the long-run.
  • Stumbling through users - actually telling SU about who is similar to your (rather than just relying on occasions when you both vote for the same content). This can be interesting when you connect with your "real" world friends!
  • Voting for content that you find through search engines. I do this a lot when I find a site at work that I wish I'd known about sooner!
  • Using the SU toolbar's search feature (All>Search...). I picked a series of topics from my interests (conveniently captured in the tag cloud on the right of this blog) and began stumbling through these topics and voting.
  • Discovering content - i.e. being the first person to vote on a page, entering uncharted territory.

The Same Old Story - Dealing with Duplicates
SU suffers from a lot of duplicate content. This is a problem with the input data, since blogs tend to be derivative. It's not uncommon to encounter the same picture on two different sites. I've even Stumbled Upon 3 blogs in a row that all claimed to offer the Best List of Photoshop Tutorials! Clearly you don't want to give duplicates the "thumbs down" - you can't tell who blogged about the content first and you wouldn't want to disagree with those other people who voted for the site (you like the content, it's the duplication that you dislike). Voting the site up would only compound the problem and will increase the chances of your getting further duplicates. So how do you respond?

There are a few ways of avoiding the duplication trap:

  • Add a comment - "Oi! You nicked this from www.SawItHereFirst.com!". This is arduous - you can't take responsibility of policing the internet!
  • Mark the content as duplicate. This is simple: Tools> Report Last Stumble> Duplicate
  • Bookmark the site instead

I tend to do the first and last. I use del.icio.us for bookmarking gargantuan lists of photoshop tutorials. I can refer back to these when the need arises and search through the list for the relevant How To (the find keyboard shortcut on Firefox "/" is brilliant!).

Update 28/11/08: The SU Toolbar v3.28 now includes the following thumbdown types that help to deal with the duplication trap!

1. Not-for me - This doesn't suit me.
2. Stumbled this before - This content was in a previous stumble.
3. Too much like this lately - Baby animals are nice, but that's enough for now.