Keep testing different models and learn something new from your dataset. There is a way of thinking that treats visualization as a magical process that prettifies the output, yet the only way we can make a meaningful graph is when we discover something meaningful during our research. People are relying on us to see the world. Out of everyone, we are the ones who need to have these interesting tests / models / questions / hypotheses at hand because if we don't know the color “red,” our story will never have “red” in it.
At 10:30 or so last night, I was on a phone call with Mike Anello and Earl, another coworker, when Mike tells us that the President is coming on TV for a special announcement in a few minutes.
I'm like, "the President of the United States?"
I searched Google News - nothing seemed to be there yet.
If you haven't noticed the lovely Lijit search wijit on the left sidebar here, please leave your feed reader and come check it out. This is a very interesting concept to me - they're taking the Google Personalized Search tools and providing a very user-friendly (and statistical) interface to the tool. A mash-up with a business model? Amazing. I've also read on their blog that they're going to start tracking comments on your sites as well and I suppose integrating that with the other statistics and search metrics you're already getting. Good times.
, I recognize the book as a brand around the idea of these "sticky" concepts.
