Comments on: discussion visualization with gephi https://darcynorman.net/2012/04/03/discussion-visualization-with-gephi/ no more band-aids Tue, 23 Aug 2016 15:14:22 +0000 hourly 1 By: dnorman https://darcynorman.net/2012/04/03/discussion-visualization-with-gephi/#comment-202738 Wed, 04 Apr 2012 20:37:43 +0000 http://www.darcynorman.net/?p=10184#comment-202738 it’s largely my fault – I exported the discussions as HTML archives, and then had to scrub the export to anonymize all content. Then I dumped that as a PDF for archival purposes. Also, the data came from Blackboard 8 and WordPress, which have different export options available, so I’d be recreating something to normalize the data and metadata anyway. Fun stuff.

]]> By: Tony Hirst https://darcynorman.net/2012/04/03/discussion-visualization-with-gephi/#comment-202736 Wed, 04 Apr 2012 17:10:29 +0000 http://www.darcynorman.net/?p=10184#comment-202736 What does the data look like in each case? Or is the manual coding actually a guilty pleasure?!

]]> By: dnorman https://darcynorman.net/2012/04/03/discussion-visualization-with-gephi/#comment-202729 Wed, 04 Apr 2012 14:21:54 +0000 http://www.darcynorman.net/?p=10184#comment-202729 I’m comparing data pulled from two completely different discussion board platforms, in different formats. I’ve had to recreate the interaction tree (setting parent-child relationships on nodes/rows) manually. Not difficult, but tedious. Thankfully, I’m only dealing with 400-ish entries…

Also, adding the Community of Inquiry coding data, and other metadata to each entry/node/row in a consistent manner has been a manual process. Which is actually OK, because I’d developed a decent workflow, and it helped me see patterns and things I could do with the data as it was entered.

]]> By: Tony Hirst https://darcynorman.net/2012/04/03/discussion-visualization-with-gephi/#comment-202726 Wed, 04 Apr 2012 08:24:27 +0000 http://www.darcynorman.net/?p=10184#comment-202726 Lots of “manual data entry”? What haven’t you been able to automate that you’d like to?

]]>