Are banks predicting divorces? Well, if there’s data to help them predict such things, they may very well use it to optimize their business.
Forbes has a couple of posts that peek into businesses use of “big data”. The first article talks about the race to build new analytics to solve challenges of large volumes of data. Here’s a snippet from Tom Groenfeldt‘s post, quoting Scott Gnau, head of research and development at Teradata:
Thought leaders in a number of industries are starting to leverage the additional analytic content from big data and combine it with what they have in large volume data stores as well. It is interesting to understand social media and consumer sentiment, but when that information is analyzed in combination with traditional consumer data it provides new, rich intelligence helping companies to identify trends and react to immediate business conditions.
According to another Forbes article, there are a number of studies that show that companies that characterize themselves as “data driven” as the best corporate performers. Now, when we’re talking “data driven”, we mean in how the company operates, not necessarily in what it produces as technology. Top performing companies are determined to use the data that they have (especially about themselves) to improve what they do and how they do it.
Also, banks are on the lookout for changes that could affect how they do business with their customers, and of course, their bottom line:
Banks, for example, worry about their customers divorcing, because divorce causes a change in credit-worthiness. No problem. They can now see a divorce coming before the couple does. All from the data.
As part of the “Computer Science or Data Science” panel at Techonomy 2011 in Tucson, AZ this week, the panel explored how data science has taken its place next to computer science as a fundamental element of information technology. New technologies are coming out seemingly every day, not only to handle big data, but to understand how to extract relevant information from the ocean of data we’re swimming in.
A company in Silicon Valley, ai-one, announced today that they have “a breakthrough method to graphically represent knowledge enables software developers to easily build intelligent agents such as Apple’s SIRI and IBM Watson”. The technology, ai-Fingerprint, is geared toward natural language programming, allowing developers to create new technologies that use natural language as input data.
Apple’s Siri and IBM’s Watson are definitely heading in the right direction for this type of technology. I just bought an iPhone 4S and I’ve tested Siri out a number of times. While Siri doesn’t get everything right (it keeps thinking my name is “Nick” when I say “Mic”), it does get more right than I expected. I was able to send texts and e-mails to people without keystrokes, and I took some notes using the voice feature, getting nearly every word correct. Pretty amazing stuff!…
Watson is the supercomputer that beat two longtime Jeopardy! champions, and it uses a technology approach that looks for the best answer for the questions being asked (or in this case, the best question for the answer being presented – it is Jeopardy! after all…). These are definitely the models that should be emulated; although, ai-one’s announcement is a press release so before we see the results, let’s chalk this up at the moment as good marketing…