As the whole world was trying to predict the winners of 2011′s Academy Awards/Oscars winners, we thought of listening to the pulse of people by measuring their sentiments and volume on various social media sites including Facebook and Twitter. We at SocialNuggets published our forecasts on the Slideshare site and now it is time to look at how, we the public, did by way of our social media conversations Vs Academy.
So this blog has three sections
- SocialNuggets Forecast Vs Academy’s Decision
- Commentary on the results
- How was our research done
SocialNuggets Forecast Vs Academy’s Decision
|Category||SocialNugget Forecast||Academy’s Decision||Comment|
|Best Supporting Actress||Hailee Steinfeld||Melissa Leo||Melissa was a distant third|
|Best Animated Film||Toy Story||Toy Story||Agreed|
|Best Foreign Film||Biutiful||In a better world||Was a distant fourth|
|Best Documentary||Exit through the gift shop||Inside job||Gaasland was the second one in social media|
|Best Supporting Actor||Christian Bale||Christian Bale||Agreed|
|Best Directing||David Fincher||Tom Hooper||David and Darren were ahead of Tom in social media|
|Best Actress||Natalie Portman||Natalie Portman||Agreed|
|Best Actor||James Franco||Colin Firth||This was the closest contest on social media|
|Best Film||The King’s Speech||The King’s Speech||Agreed|
Commentary on the results
It is clear that Academy doesn’t always pick the most popular public opinion as expressed in social media. Our record of prediction was about 50%. So, can you really use the social media conversations to predict the winners? The popular opinion does coincide with the commercial success of the movie. Perhaps, someday Academy’s opinions will align with the general public but until that time we will continue to learn from this experience and modify our algorithms and will be back next year with more.
How was our research done
Using SocialNuggets’ technology of focused harvesting and text analytics and sentiment engine, we started mining social media conversations for the last few days not just for mentions but also for sentiments. We settled on only 9 award categories as there were not enough meaningful conversations to predict in other categories until 2 days ago. Using our engine, we came up with net perception score of each of the mentions and then multiplied the % of mentions with Serendio Net perception score to get a final score which was published in our slide share discussions.