I spent the afternoon testing internal cabling looking for a transient error. It now looks like we have the bit error rate (BER) under control, BER both pre and post test are now 1.0E-9. The FEC reference in the image refers to forward error correction (FEC) first introduced by American mathematician Richard Hamming who pioneered this field in the 1940′s.

According to wikipedia and what I remember from the Tanenbaum and Stevens books: FEC gives the receiver an ability to correct errors without needing a reverse channel to request retransmission of data, but this advantage is at the cost of a fixed higher forward channel bandwidth. FEC is therefore applied in situations where retransmissions are relatively costly, or impossible such as when broadcasting to multiple receivers. This is the case with cable data connections.

What was old is new again.

(See also: Richard Hamming)

While exploring my Twitter social graph I ran across Pierre Lindenbaum’s twittergraph Java tool for mining Twitter networks. It produces a GEXF file in XML citing those relationships. The command-line looks like this:

java -jar  twittergraph.jar -o twittergraph.gexf 14087619

As Pierre describes in his post, his tool doesn’t use the OAuth API, so it has to wait for a few minutes and retry the connection every time it reaches the twitter API quota of 150 requests per hour. Given my id number this script ran all night. Once the GEXF file is created you can explore the connections and clusters with Gephi 0.7 alpha. I’ve used the Force Atlas method to sort edges and nodes. In preview mode, with 30% of the edges rendered, clusters are clearly visible around Life Sciences, Visualization, and HPC feeds.

Like Pierre, I used zoom.it to embed a PDF version of the master network document in this post. Here is a view of my complete network:

Data privacy issues aside, I’ve come to like zoom.it for viewing large graphs as embedded web documents.

Other methods of mining the social web and mapping professional relationships might include projecting LinkedIn contacts onto Google Earth. This recent O’Reilly article does just that.

(See also: Yokofakun: Visualizing my Twitter network)
(See also: github: TwitterGraph.java)
(See also: gephi.org)
(See also: zoom.it: hosted twitter graph for tkunau)
(See also: How to Map your Professional Network with Google Earth)

I’ve been selected to speak at MinneWebCon in St. Paul, MN on April 11, 2011. Here are my session details:

More than 1000 Words: Scripts and Techniques for Interactive Graphics

This session will explore scripts and techniques for putting interactive graphics in web sites. Methods for displaying complex data sets on cross-platform browsers require adherence to broadly accepted standards; today those standards include HTML5 and JavaScript. The HTML5 canvas element provides a native JavaScript drawing API, allowing us to dynamically draw bitmap images on a page. Current and emerging JavaScript tools will be covered in this presentation including InfoViz, Protovis, Raphaël, Processing.js, and the jQuery Visualize plugin.

(See also: MinneWebCon)