Category Archives: data

More website ranking fun

Following on from this post about website ranking standards, I have found a new toy.

dnScoop provides another wideranging rating for your website, taking into account a stack of search metrics. The output is a value for your site in $s. Apparently the blog you are reading is worth $121. I’m rich!

On one level it’s a shameless attempt to earn textlinkads referal revenue, but good luck to them. Assuming doesn’t eat the code (it did!), here is an example of the output.

My Site is worth $121

How much is your site worth?

Visualisation automation for work and movies

One of the biggest challenges I face on a regular basis is how to convey relatively simple financial projections to the number blind on my board. Graphs, charts, slaps in the face, all fall on deaf ears. Eventually I’ll hit on something that makes them go ‘oooo okay yeah I see what you’re saying’. Eventually. Using Excel this can take forever.

Take a less straightforward dataset, say the behaviour of a set of users of one of our systems which leads me to propose a certain change to a process. How do you convey THAT without a week of ‘I don’t get it’.

New toy time.

The ‘many eyes’ project has some nice examples. The holy grail of visualisation has to be an automated system which you throw a dataset at, and it proposes an endless sequence of possible visualisations. I have a big screen, show me lots of options. The following is a very wordy example:

holy grail

Words used in Monty Python and the Holy Grail. Obviously. It’s a tag cloud. Big deal.  Hitting the ‘visualise using this dataset’ button brings up a stackof options. I can think of a dozen nice ones that aren’t already there. One frustration is that it doesn’t filter out the plots that the data set wouldn’t support. Even better would be a wizard to give you options for creating the data elements that you need to make a certain plot. For example I could define an X axis as the Scene and the Y axis as the distance through a scene. I could then visualise the data above in relation to how far through the movie the words were. Bottom left are words that appear early in the movie and the scenes. Top right are closer to the end of the film or scene. I’d guess the punchline words will be further to the right and top.

Plot this for a thousand comedies and you should certainly see if the punchline zone top right has interesting language in it, or just terms of ridicule. I’m off on the hunt for more projects, but if you know of any drop me a note. Link (via)