IBM heralds new era of social network analytics

IBM is working on sentiment analysis and stream analytics

Companies are increasingly eager to analyse information from social networking sites, but are being overwhelmed by the volume of data, according to IBM.

The firm said at a press event at its Armonk headquarters this week that customers regularly ask how they can collect and analyse the huge amounts of information about their products and services now found on sites like Facebook and Twitter.

The main barrier is the sheer volume of information being generated. IBM pointed out that seven terabytes of data is uploaded to Twitter alone every day, but claimed that its analytics systems can narrow this down.

"We can hook up to the Twitter feed of any company and pull out 10 minutes of information," said Rod Smith, vice president of emerging technologies at IBM.

"Using a sentiment analysis, comments can be put into context to give an idea of what people are saying."

However, the way language is used can make analysis difficult. Words have to be weighed carefully, and context is very important. Smith gave the example of the word 'wow', which could be positive but could also be negative as in: 'Wow! This product sucks.'

Businesses are also interested to see how far they can push the boundaries of user privacy in this area, investigating how tracking customers can benefit their brands.

Another emerging area is stream analytics, which can be seen as a "feeder to an analytic system", explained Smith.