The Data Source Handbook by Pete Warden provides a concise and handy guide to some of the main sources of public data accessible on the web today. It’s a very short book of 40 pages. This in itself does not stand against the book. These sources are rapidly changing and compiling and committing an exhaustive survey to a printed volume would damn it to almost instant obsolescence. It would also prevent any treatment of individual datasources in any useful detail.
In Mining the Social Web Matthew A Russell offers to instruct in identifying social connections, trends in discussion and locations by tapping into social media data. He succeeds in spades. This fast-paced and rich handbook jumps right into the fray and provides an immediate and useful exercise in accessing the Twitter API using python and doing a very quick visualisation of trending subjects. I was hooked and greedily and immediately consumed a few more of his lessons. His approach is to go direct to real world applications of why you’d want to mine data from social media such as Twitter, Buzz, Facebook and utilise other freely available tools such as Google Maps to look for patterns and present solid research findings.
The folks at Many Eyes recently introduced their new comparison cloud tool. Basically, it lets you visualise two fragments of text displaying word frequency for each in the same cloud. It’s an interesting addition to the more familiar word cloud. 
Using a standard word cloud you get a matrix of words with relative size, weight or colour highlighting frequency in a selected text. This quickly allows you to visually perceive an author or speaker’s emphasis on a particular theme or style of writing or speaking. With Many Eyes hybrid tool, words which occur in both text are abutted. You can now visually compare two texts from the same author for similar empahsis or quickly determine a difference between texts. In the example presented at Many Eyes, they compare the US presidential State of the Union addresses from 2002 and 2003. In this example they note the less frequent mention of Afghanistan and the increase in mention of Saddam. Whether this allows one to conclude a change in policy or not, it does demonstrate the use of the tool for provoking questions for further exploration.
On Saturday, the Ontario government officially announced how much funding each university in Ontario is to receive for maintenance and renewal of facilities. I just happened to see announcements from a few institutions appear simultaneously in my RSS reader and was struck by the rather different ways in which they presented this news.
A year ago I wrote a recipe for the TAPoR project to demonstrate a way for historians to utilize text analysis tools to plumb historical data from Google. In the recipe a user aggregated search results from Google and used the TAPoR DateFinder tool to rapidly construct a chronology. This rather basic operation has now been automated by the folks at Google labs. Now, with the simple addition of two words in your search request you can choose to view the familiar text search results in two exciting additional contexts, temporal and spatial. The new Google Timeline and Map views is a powerful but simple tool for historians and others as well. Read the rest of this entry »
I admit it…I had no idea what a contronym was until I saw this fascinating list of examples. A contronym is a word which its own antonym. Get it. Maybe not. The word custom for example means both usual and also special . I like that one! Check out this web page where there is a wonderful smattering of words that meet this criteria — ordinary everyday words which we use and possibly have never thought twice about. English can be such a funny language.
This is a just really cool. Knutz.net Thomas Broome has a series of pictures drawn entirely with words…really. Interior landscapes composed of chandeliers drawn only using the word chandelier. Wonderful little details to appreciate.
Clever lads have run the CES address of Bill Gates and the Macworld Keynote by Steve Jobs through a variety of text analysis tools to get an idea of why one has greater impact than the other. The article demonstrates that there is a huge difference in the complexity of the message. Jobs delivers short, easily comprehended sentences, where Gates tends to be using longer sentences, with more complex language. The word clouds generated from the speech are not that different in terms of focus. Both featured most frequent references to the products being featured. Interestingly this contrasted with Michael Dell’s CES presentation which was seemingly used much more ambiguous language with less direct reference to particular products. There’s also a slider-based version linked to the article that offers an alternative way to view the clouds. Unfortunately unless you use the arrow keys (i.e. read the small print) it seems next to impossible to click on the magic spot to get Gates cloud displayed.
This exercise begs the question of magic however, and whether it is merely the message and not thew actual technology being presented that enthralls the audience. One would expect that the concept of the iPhone itself may actually be more appreciable than Windows Vista and Michael Dell simply didn’t talk as much about products because he didn’t have any exciting new product to introduce. Nonetheless, a fun little intellectual exercise.
Gates in fact doesn’t seem to have always had the product focus that he does now. There is a word cloud timeline of his communications and it is only recently that products have bgun to experience high frequency of reference.
Everyone knows that the value is in the network. SNA is a wonderful tool for academic and I am using it to map my local webs of commerce. The folks at Spoke however are doing this one step better. They have the typically enormous and touted list of key decision makers and influencers at companies around the world. Nothing short of a big Spam list there. However, on joining the network, you contribute your own contact list. Again nothing revolutionary in that…but here’s where it gets interesting. The little client that harvests your contacts for Spoke also measures how connected (inbetweeness in SNA-speak) you are based on frequency and nature of contact based on your email history. Sure its not flawless, but when you overlay this with all the other participants they are building one mega web and are creating a potentially rich map of influence flows. It raises some serious privacy and trust issues, but it is clearly pushing the envelope one step beyond. Many CRM apps are out there trying to build similar webs in an automated fashion, but generally require huge rejigging and manual creation of hierarchical relationship by thew user. Few actually automate the process, let alone start to weight the results through contactedness (not connectedness) mining. Intriguing.