R help - Aug 2009 - A request for pointers on how to analyse Twitter with R ?

Hello dear R members.

I just came across this website:
http://giladlotan.org/viz/iranelection/

Here what it is about:
> ReTweet Revolution is a visual exploration of the most popular conversation
> threads that were passed amongst Twitter users at the time of the events
> following the Iranian elections in June of 2009.

It uses an algorithm described here:
http://giladlotan.org/viz/iranelection/methodology.html
> A very simple rundown of my analysis process:
> 1. Public Timeline Polling - During the period of the #iranelection events,
> my I would poll Twitter's public timeline, looking for keywords such as
>
"#iranelection","#gr88","ahmadinejad","mousavi".
> 2. Text Matching Algorithm - as new tweets entered my database, making
> matches between those that belong to the same conversational thread.
> 3. ReTweet analysis - mapping out the network structure by which messages
> spread between users.
> 4. Text differentiation analysis - looking at syntactical differences
> between tweets from within each thread. This highlights the "broken
> telephone" aspect of ReTweeting, where every user can choose to change
and
> add to the message as it is passed onwards.
> *more to come*


And I would love to be able to do something even similar to what he does
(simply because it is cool).

They way I see it it requires several steps into making this happen,
including:
1) a way to pull and store data from twitter
2) a way to extract the data from the database
3) tools to analyse and visualize the data (this, I imagine, can be done
using the "tm" package, the rest I don't know how to approach)


Can any one give interesting pointers on this one ?

Thanks,
Tal




-- 
----------------------------------------------


My contact information:
Tal Galili
Phone number: 972-50-3373767
FaceBook: Tal Galili
My Blogs:
http://www.r-statistics.com/
http://www.talgalili.com
http://www.biostatistics.co.il

	[[alternative HTML version deleted]]