Blog mining

| April 11, 2010 | 0 Comments

The Economist: The vast majority of the millions of blog entries written each day sink into obscurity, read only by their authors, but thanks to research from the University of Southern California‚Äôs Institute for Creative Technologies, they may soon have at least one avid reader – a computer. Researchers are hoping to one day ‘data mine’ blogs from around the world for information, in the same way Google Analytics looks for interesting patterns in search terms. But to do that, they need computers to be able to analyse the human experience.

So they’ve been teaching computers about cause and effect – causality – to enable them to follow the narrative in a blog and draw out useful information. This approach could provide interesting data on trends in human behaviour, such as drug use, racial tension, new products or film preferences, The Economist reports. Read more here.

Leave a Reply