This is by way of slamming down a reference to Avinash Kaushik's blog so you can keep track of it. He is a guru of web analytics. What makes him interesting is that he is good enough to step away from clicks and to propose integrating a trinity of data sources - clickstream, business outcomes and surprise surprise customer research. I want you to take a look at his blog post on measuring social media which is notoriously easy to do and difficult to do well. He is working off piste but his lucidity is really helpful. What I got particular excited by was tweetpsych. A tool which aggregates a ton of tweets and then generates a psychological profile. Kaushik dismisses sentiment analysis pos/neg tweet classification as twaddle. He's looking for analytical tools which can pick up irony, emotion, making authoritative statements - content of that kind. Which tweetpsych appears to do. I commend it to you.
written on 11-Dec-2009
Paul Hutchings [http://kindler.posterous.com] says:
I’m still scratching my head about whether social media is best for qualitative or quantitiatve type analysis, or even if terms like that are still relevant. There is a mass of data out there that could be quantified in some way and we may well get some qual insights out of that. But do people really write/ type stuff in a public environment that really gets at what they are thinking and feeling? They certainly don't verbalise it unless they get asked the right questions and some bright spark makes sense of it without taking it at face value.