We would like to build an interactive visualization tool that allows
exploration into how people are interacting with a website. This
visualization allows you to see how users on average navigate through
a website and then to do some pivoting on various user attributes and
look at trends over time. The pivoting could, for example, show how
people from US vs non-US click-behavior varies. The trending could
demonstrate that users are exiting at a higher rate after a certain
point in time. The work uses massive clickstream data sets from
websites. The visualization I imagine is a directed graph that shows
click-through rates and exit rates from the site. Here is something
similar that IBM’s Many Eyes built using processing:
http://manyeyes.alphaworks.ibm.com/manyeyes/visualizations/word-tree-
monica-goodling
I’ve built the components that get the data into a format that can be
used by some visualization software. I’m currently trying to build a
prototype in Processing but it would be great to get someone involved
who has other ideas to make the interaction compelling and easily
digestible to a user of this tool. I already have something working
using python scripts and GraphViz but it is not interactive.
Pay is flexible depending on experience and much of the work can be
done remotely and independently. Please contact me at
jrgoldman@gmail.com if you are interested.
—
