Oscar Gets Social

 
 
By Samuel Greengard  |  Posted 2012-02-29
 
 
 

This year’s Oscars provided the usual array of upsets and surprises. However, a social media project sponsored by IBM, The University of Southern California Annenberg Innovation Lab and the Los Angeles Times tracked Twitter sentiment before, during and after the Academy Awards and found that, in many instances, the pubic didn’t see eye-to-eye with the way Academy members voted.

Although George Clooney went home empty handed for his performance in the “The Descendants,” he earned “The People’s Oscar,” based on positive-to-negative sentiment from Twitter. For best picture, “The Help” ranked highest for positive sentiment, followed by “Midnight in Paris.” For best actor, Brad Pitt (“Moneyball”) ranked highest for positive sentiment, followed by Jean Dujardin (“The Artist”). Twitter sentiment also revealed disappointment that “The Girl with the Dragon Tattoo” wasn’t nominated.

The analytics technology—which has been previously been used for the Super Bowl, World Series and in the retailing industry—relies on semantic and linguistic analysis to identify positive, negative and neutral sentiments, distinguish irony, and even apply 'machine learning’ to determine which tweets are merely background noise and those that make an impact. The software offers insights that aren’t possible with traditional media. USC researchers and students worked with IBM to develop the analytics capabilities.

Rod Smith, IBM Fellow and vice president of Emerging Internet Technologies, says that the project has value far beyond the Oscars or NFL. Fields as diverse as retail and journalism can benefit by tuning out social media “noise” and identifying key issues that frame everything from customer segments to marketing strategies. “It represents a real paradigm shift in how businesses understand, interact with and uncover actionable insights,” he says.

Understanding customer sentiment is increasingly critical, whether it involves a movie studio shaping a marketing campaign for a film or a retailer understanding how the public views a product or its brand, says Jonathan Taplin, director of USC’s Annenberg Innovation Lab. What’s more, the task is growing more difficult as the signal-to-noise ratio for social media feeds rises. IDC predicts that a 29-fold increase in data volume will take place over the next decade. “Businesses that don’t engage in the social media conversation are operating at a distinct disadvantage,” Taplin concludes.

For a look at how the public view squares with Tinseltown for the Oscars, visit the Los Angeles Times Senti-meter.