Session 3C: Annotations in Viz & D3

What is everybody interested in round table:

creating full conversation, tech documentation, user communicate, storytelling, user annotation, notes over time, annotations own dataset, learning viz, as a starting point for exploration

let's look at some examples:

• NYT example
• Many Eyes
• Comment Space from Heer's lab - explore data, take a snapshot & add a comment, many people do this
• storytellingwithdata.com

As visualization evolves over time - the diff. What's normal and what's different from normal. What are the highlights that are important because they are different.

Tamara (paraphrased): you don't want to get to the point where the users are doing search for you - annotation aids the search. Ex: if you're letting the user hunt & peck, you're asking them to search for insights vs surfacing them.

As the author of the viz, you should tell your story. But, that others can suggest their interpretation/their story.

Ex: in the kindle you can see the most popular highlighted sections in the book. Can change how you experience the book by seeing what others see as important.

With D3 - bind an annotation to a visual element or data point. Or bind to a state (but how do you define that). Or - this is the grid, and annotation shows up here.

Great talk by Adam Pierce at Bloomberg (not sure if recorded). Render visualization as SVG, go to illustrator, add annotation exactly where they want it, then put back in.

Annotation as a challenge around getting the right collaboration between automation & humans.

Showing you what others have done, could be interesting. In the pixel space... or the dataspace.

You draw it - show others have same misconceptions as you (or different).

Veritasium example [zan to do: post link & explanation]

Look to things like Google Maps, Open Street View, etc. What do other's do?

Using people to derive information. Example: recaptia & a library example from Susie. The annotations people write can become the real data.

Example: Rap Genius. Many companies have tried to crowd source annotations - this is an interesting example. Rap Genius: it's an interpretation of an abstract layer. Some will be way off-base. There is a level of accuracy, even if it's not totally right/wrong. There is a gatekeeper. --> Is it necessary to have a hierarchy in the annotation community? Ex: Stack Overflow & Wikipedia.