about pre-collected twitter dataset

Sep 6, 2013 at 6:33 PM
Hi,

I already pre-processed the data from twitter.
The data contains ID with all contents that ID tweets, and date.
(e.g.
815299, I'm at Bleeding Heart Bakery (1955 W. Belmont Ave, Damen Ave, Chicago). , 2/19/2010 8:43)

In order to import this data to NodeXL, do i need to manually extract nodes ? or
Is there any way that I can extract nodes automatically, using node XL?
Thank you in advance!

Jinie
Sep 6, 2013 at 8:08 PM
Hello, Jinie:

NodeXL knows how to pull "mentioned" and "replied-to" usernames from a set of tweets and then create a network using those usernames and the relationships among them. However, it can do that only in the context of the Import from Twitter Search Network feature, where NodeXL gets the tweets directly from Twitter, along with a bunch of metadata about the tweets and their tweeters. It cannot process a set of tweets obtained elsewhere.

So you will need to do that processing yourself. I don't know how you've done the processing so far, but I'll point you to the code that NodeXL uses to pull usernames out of tweets, in case it's of any use. It's in C# and it uses a couple of regular expressions to do the job.

TwitterStatusTextParser.cs

http://nodexl.codeplex.com/SourceControl/latest#Common/SocialNetwork/Twitter/TwitterStatusTextParser.cs

-- Tony