NodeXL and DataSift

May 26, 2014 at 12:25 PM
Dear fellow-analysts,

Recently I've been working on a research project involving the use of twitter during crisis response. During our exercise we have had many volunteers who tweeted from their own account using a pre-assinged profile (role).

Since I know that NodeXL is depending on Twitter's API to provide tweets and cannot guarnatee a full set of tweets, I've taken a DataSift account and collected the relevant tweets during the exercise using their record function.

My goal is now to import those tweets in the correct format into the NodeXL spreadsheet so I can use your great tool. So far I've figured out that I need to make seperate entries for each posisble interaction and enter that in the Edges sheets, Vertrices should be calculated from this sheet if I'm not mistaken in NodeXL or possible augmented with profile information.

If the experts here want to help me figure out where which data should go, I'll be happy to rewrite this into a tutorial so others can do the same. Perhaps we could work on import function? DataSift allowed me a trial version to collect tweets so it might be usefull for others as well in the future.

In short: where should which data? Is this documented somewhere? Thank you for the support!
May 27, 2014 at 5:41 PM
Edited May 27, 2014 at 5:42 PM
Hi kmeesters,

Is the data that you got from DataSift in CSV or JSON? Either way, you're going to need to write some code to parse the data to extract the relationships before importing your data into NodeXL. Check out this thread that I started a little while ago where Tony gave me some suggestions on how to extract a Twitter network from JSON data. Import JSON data from Twitter Streaming API to build a network from