Visualizing Domains Projects

One of the big challenges in running a Domains project, and part of my feeling of being adrift at sea, is the highlighting of particularly good work from users and the intelligible visualization of the broader activity from all users.

At OU, we have a couple of sites for this purpose. The Activity site shows the most recent blog posts from all sites that are capable of being read by Feeds WordPress. You can get a preview of each blog post and then click on the link to go to that web site. Each week during the school year, we put out our own blog post featuring the top 5 or 6 posts from this weeks Activity feed.

A second project, called Sites, is a filtered list of all of the web apps currently running on all four of the OU servers. This is really nice for finding the links for all sites running Vanilla Forums or dokuwiki or even more commonly used apps like Drupal and Omeka. However, if you filter for WordPress, you still get a simple list of the 3400 or so sites using WordPress.

At Domains17, Marie Selvanadin, Tom Woodward, Yianna Vovides talked about the answer that they are currently developing for Georgetown. Their design starts with a search bar that searches across all blogs. A second piece is called themes. Another option implements the TimelineJS library to visualize posts by a given user or in a given theme along a timeline.

The great @twoodwar and his Google spreadsheets. #DOMAINS17 pic.twitter.com/9FT5L7FfY6

— Jim Groom (@jimgroom) June 5, 2017

Currently, Tom is running the first iteration of this visualization of the Georgetown domains off of a Google Spreadsheet, as he is want to do. He then generates a front page with the basic metadata and screenshots of the sites. For each of these sites, there is a dynamically generated page with the text of the various posts, word counts, and a charts visualization of the word counts. You can also see the timelineJS by category for each site.

The next phase of research is to dig into the visualization of community sites. How do we include closer analysis of Drupal, Omeka, and other apps along with html sites, rather than just WordPress sites? What types of questions can/should we ask? What are the ethical questions around mining this data?