Follow Datanami:
May 21, 2014

‘Social Physics’ Harnesses Big Data to Predict Human Behavior

A widely respected big data guru is positing the notion that the explosion of personal data being generated by sensors and other networked devices is creating a kind of “social physics” that can propel a flow of ideas that transforms those ideas into human behavior.

That’s the thesis of a new book, Social Physics: How Good Ideas Spread-The Lessons from the New Science, by Alex Pentland, a computational social scientist at Massachusetts Institute of Technology and director of human dynamics research at the MIT Media Lab. Pentland is a pioneer in the field of social physics, which focuses on leveraging big data to figure out how human social networks spread ideas and transform human behavior.

Pentland has combined his earlier ideas on social physics with the rise of big data to examine what these enormous new data sets tell us about why humans do what they do. Pentland’s research draws on the “digital bread crumbs” left behind by smartphones, other mobile devices and, to a lesser extent, Internet habits.

According to his publisher, Pentland and his team found that they could use this personalized big data to study “patterns of information exchange in a social network without any knowledge of the actual content of the information and predict with stunning accuracy how productive and effective that network is….” Those predictions include networks ranging in size from a business to an entire city.

“At every level of interaction, from small groups to large cities, social networks can be tuned to increase exploration and engagement, thus vastly improving idea flow,” asserted the editors in an overview of Pentland’s latest book.

These trends in big data analytics and social physics research are being driven in large part by the proliferation of mobile devices and sensors that are generating enormous data sets. Add to that, devices themselves are being connected via ad hoc networks that have come to be called the “Internet of Things.”

The result is a flood of unstructured data that researchers have just now began to probe for insights into things like consumer behavior and personalize health care.

As a leading big data researcher, Pentland has proposed new concepts like “reality mining” as part of his efforts to collect and interpret big data sets. In the reality mining example, researchers seek to figure out what people are actually doing, not what they are saying.

Pentland recently told the New York Times that tracking a person’s daily movements via smartphone GPS signals and credit-card transactions is far more revealing than, say, Web browsing habits or social media comments. What people do is deemed critical by social physics researchers, not what they say or think, Pentland asserts.

Pentland also serves as an advisor to the World Economic Forum on big data and privacy. In that role, he has proposed a “New Deal on Data” that would among other things allow individuals to own their data and control how it is used.

The proposal stems from a sort of “big data backlash” generated by concerns about data privacy. Critics warn, for example, that data brokers can gain access to and sell personal medical data without a consumer’s knowledge.

“I see the public good that can come from using this data,” Pentland told the Times. “That’s why I got involved in this privacy stuff.”

Related Items:

Personal Data Is Not Forever, EU Court Says

‘Urban Data’ to Shape NYC Development Project

Open Data Initiative Looks to Seed New Markets