IBM Targets the Future of Social Media Analytics
This is from back in April, 2011 but thought it was worthy of a note. The post reads in part:
The new product, called Cognos Consumer Insight, is built upon IBM’s Cognos business intelligence technology along with Hadoop to process the piles of unstructured social media data. According to Deepak Advani, IBM’s VP of predictive analytics, there’s a lot of value in performing text analytics on data derived from Twitter, Facebook and other social forums to determine how companies or their products are faring among consumers. Cognos lets customers view sentiment levels over time to determine how efforts are working, he added, and skilled analysts can augment their Cognos Consumer Insight usage with IBM’s SPSS product to bring predictive analytics into the mix.
The partnership with Yale is designed to address the current dearth of analytic skills among business leaders, Advani said. Although the program will involve training on analytics technologies, Advani explained that business people still need some grounding in analytic theory and thinking rather than just knowing how to use a particular piece of software. “I think the primary goal is for students to learn analytically,” he said, which will help know which technology to put to work on what data, and how.
Within many organizations, he added, the main problem is that they’re not using analytics at the point of decision or across all their business processes. Advani says partnerships like those with Yale will help instill the thought process of using mathematical algorithms instead of gut feelings.
I was with them up to the point that it says: “….instill the thought process of using mathematical algorithms instead of gut feelings.”
I don’t take “analytical thinking” to be limited to mathematical algorithms.
Moreover, we have been down this road before, when Jack Kennedy was president and Robert McNamara was Secretary of Defense. Operations analysis they called it back then. Should be able to determine, mathematically, how much equipment was needed at any particular location and didn’t need to ask local “gut” opinions about it. True, some bases don’t need snow plows every year, but when planes are trying to land, they are very nice to have.
If you object that is an abuse of operations theory I would have to concede you are correct, but abused it was on a regular basis.
I suspect the program will be a very good one along with the software. My only caution is really on any analytical technique that gives an answer at variance with years of experience in a trade. At least a reason to pause to ask why?