Tag Archives: nodexl

26 Feb

Khojaly hashtag analysis

There is currently a campaign “Justice for Khojaly.” I wish that I could link to wikipedia or some other source with confidence, but this issue is so fraught with tension, it is very difficult to find a neutral source.

With that being said, here’s the wikipedia.

So lots of Azerbaijanis have been tweeting about this. So much that there were over 1500 tweets in the last few hours alone.

Here’s the analysis.

nodexl

You can see on the far left a lot of people who are not networked with anyone else but tweeted the word Khojaly. But then you can also see a number of clusters of people who do follow ecah other.

Top 10 Vertices, Ranked by Betweenness Centrality — so this means, who are the MOST NETWORKED people in the analysis:
ceyhunosmanli
AzNewsNetwork
nurana21
HDNER
GNLZYNALOVA
fuadshahbazov
elmanabdullayev
FidanKerimova3
Ziya_Meral
Elnur_Z

Top Tweeters in Entire Graph — this is who tweeted the most:
fhhknews
PeedroDionisio
kardeleniye
erolmaras
MankeErich
googlyfish_ca
news_az
todayaz
barbarosturpcu
Alibey_Aze

Honestly, I don’t know anything about any of these accounts so I can’t contextualize much, but I figured that people would be interested.

24 Feb

Please explain these knots of string on your blog in plain English

I have discussed this before, but from a more abstract perspective… what is all this social network analysis stuff?

A network is a collection of things that are connected to one another.

By looking at the network, we can see the relationships between members – clusters of people, who is important, who is connected, how the relationships change over time. Some of this stuff is intuitive – like popularity or influence, but network analysis can go deeper.  For example, how many of your friends know each other? Are there network members that are bridges between groups?

We visualize the network and use symbols and colors to show patterns.

—

Social media is cool because it gives us a TON of data – not only content, but conections. We can collect, analyze, visualize, and make inferences on these connections.

—

Social network analysis tl;dr:

Nodes are actors.
Edges are connections.

Centrality is the number of direct connections that individuals have with others in the group.

Cohension is the ease with which a network can connect – short paths between nodes.

Density is the number of connections.

Betweenness is the shortest path between two nodes.

We can also know attributes of nodes – when did they join Twitter? How old are they? And map this onto the visualization.

23 Feb

#armvote13 -> #barevolution

So now that some real stuff is happening in Armenia, we have a new hashtag! #barevolution or Բարեւոլյուշըն in Armenian – so this is a play on words. Raffi say “Barev, Hayastan” to the crowd – meaning “Hello, Armenia.” Moreover, the word “arev” means sun.

I’m not really sure if this has totally caught on as a hashtag yet, but it might.

So there is a cute logo for this.

nodexl

But this makes hashtag analysis a bit more difficult. While Eastern Armenians (those in the Republic of Armenia) would say Barev, Western Armenians would say Parev. And the way that “ev” is spelled is different.

So, here are the hashtag analyses for Feb 23, 10am Yerevan time.

#armvote13 – will post later

barevolution

nodexl

parevolution
Anything containing “barev”