IT leaders at Computing's Big Data Summit 2013 yesterday agreed that social media analysis could benefit them internally as well as externally, and could be the next step for them in order to get the most out of their big data tools.
Tony Scott, global enterprise architect at Arup, said the design and engineering firm has done a lot of work on smart cities using what he called "reverse social engineering", to predict the London riots, for example.
Further reading
Big Data is more than just the datacentre, says Intel Be wary of companies rebranding to offer 'big data solutions', says FT CIO Big Data will be a game changer in fighting diseases, says AstraZeneca Analysis: Is there a need for social analytics in the enterprise?
But now, he said, the company was installing social enterprise tool Yammer in order to apply sentiment analysis internally.
Although the firm has not used internal sentiment analysis thus far, his thoughts struck a chord with his fellow panellists at the summit.
Financial Times CIO Christina Scott said that social media was hugely disruptive to news organisations.
"We can listen to what people say about the FT and particular articles, and journalists, and feed sentiment analysis to customer service agents to gauge how happy they are, or the reasons a subscriber is looking to leave," she said.
While Scott said that the FT was in the early stages of using social media analytics, she was encouraged by the idea of using internal social media analysis, as was Martyn Croft, CIO at the Salvation Army.
The Salvation Army currently uses some external sentiment analysis, Croft explained.
"It is about understanding the support and measuring the support to make sure the right people are behind you. Sentiment analysis is of great importance; we aren't there yet but a lot of other charities certainly are," he said.
Meanwhile, Transport for London's (TfL) head of analytics, Lauren Sager Weinstein, said that the organisation has not yet used social media analysis, but that it does have its own social media accounts.
"We have social media accounts that inform commuters about disruptions; we are able to take analytical information and feed it out to our social media accounts to push the information out as quickly as possible - that is not our analytics piece, but it is how you link [big data and social media] together and make it with some sentiment," she said.
Weinstein added that TfL is sometimes made aware of problems by social media before TfL systems have acknowledged a problem.
In terms of big data, Weinstein said TfL wants to make use of its passenger data, as opposed to running costly surveys filled out by a handful of commuters.
"It is still early days for us and there are things to explore. There is a lot of potential to look at the data of our Oyster cards. We want to pull together data on how passengers are travelling across the network, once we have done that we can move away from surveys and ask how we structure that data," she said.
"Surveys can have different results depending on the time people have to answer them and how the questions are asked, but when you use data of Oyster taps it is very clear. Of course this has to be looked at in a broader context as it is a complex network, but we are confident that information gathered can lead to the right kind of information," she added.
