Oscars 2015 on Twitter: Identifying Trending Topics
This is the first post of a 3-part series a couple friends and I wrote about analyzing the 2015 Oscars on Twitter.
As technologists and amateur sociologists, we love watching how conversations spread and trends emerge, and dedicated a sunny Sunday afternoon to observing the Twitter stream around this year’s Oscars.
Fueled by prosecco and tapas, we spread out our Macbooks and got to work. How we collected data from Twitter firehose, stored into Google BigQuery and analyzed it via Dataflow are described in this other more technical post.
While tweets were flowing, we watched as hashtags surfaced in popularity and began assigning them to categories, first by taking our best guesses based on noteworthy Oscar moments, then by validating with samples of data.
Still in our test phase, we could then reduce these macro categories to 5 smaller dominant conversation themes:
- “Top” which was a swatch of all the top trending topics on Twitter
- “Red carpet” which highlighted all the gossip on the red carpet fit to gab about
- “Actors” which showcased the best of the best, and the best dressed
- “Films” which covered the winningest (or most tweeted about) films of the year, and of course:
- “Noteworthy” which captured performances, oops moments, oddities, and memorable moments (such as Gaga’s dishwashing gloves and Reese withoutherspoon)
We then created a front-end application to display the data and the results of our analysis. The five themes can be clicked on to display a graph of the most relevant hashtags.
The graph is interactive, it shows the hashtag count per minute (over a sliding window of 10 minutes) and is updated on 5 second intervals. It’s got a slider that you can use to adjust the time dimension, and a hover-over that shows you the specific trending tweets at any given moment. You’ll also see the relationships amongst them, what happens to the others when one surges or falls.
We’ve enjoyed crunching data during the Oscars and finding correlations between the highlights and the spikes in our graph, and hope you had an enjoyable Oscars evening viewing. What was your favorite moment? Can you find it in our data? Let us know!
