An Interview with Graham Mackintosh
Cesar.Liao 270005NH7A Visits (952)
During our trip to the E2 Innovation Conference we got the opportunity to speak with Graham Mackintosh, the program executive for IBM’s Social Media Analytics department. He gave a very interesting session on the topic of using data gathered from social media networks and using it different applications. These applications include things such as recognizing flu patterns or whether there are areas around the city that have unusual amount of traffic. Currently it is being used to analysis possible customers determining their interesting and providing the right kind of advertisement to meet that consumer base.
We also got a chance to sit down with Mr. Mackintosh and ask him some questions:
1) How does IBM’s social media analytics compare to Google +?
Google + does some keyword matching while IBM’s system breaks down the sentence and tries to understand the language that is being communicated.
2) Is this system real-time or not?
It is both: there is a real-time element that streams terabytes of information per second and there is a more in-depth analysis that is offline that usually takes 24 hours to grind through.
3) How do you get analytics on people who don’t open up their data to you? (Ex. China)
We support some Chinese analytics but what we are able to access is constrained: we do not take information that isn’t public.
4) How do you keep this technology from being misused? (Ex. Government using it to spy on citizens)
There are two ways of getting this technology, you can either buy it from IBM as a licensed product in which case IBM gets to choose who they sell it to or you can get it as a software as a service where IBM gets to run the infrastructure.
5) Do you see this technology as being used to predict crimes? (Ex. Minority Report)
IBM is working with police already in cities like Vancouver to use this statistic analysis technology as a predictive crime system. It isn’t exactly like Minority Report but they can determine areas where there are high crimes or where crimes are likely to happen.
We would like to take this opportunity to thank Graham Mackintosh for taking the time to speak with us and answer our questions, his Social Media Analytics session was not only informative but exciting as well. We can’t wait to see what IBM does with this technology next.
Cesar Liao is a Senior-Year student at San Jose State University.