For the past few days, I have been looking at tweets about the Golden Globes. I collected the tweets using the amazing rtweet package. I then used substring search functions to look for the names and handles of prominent actors and winners (Variety’s article about Golden Globe social media trends was very helpful in this regard).
To create this line graph, I constructed a dataset with counts of tweets using #goldenglobe, as well as this hashtag with the specific users/movie mentioned below (I used their first/celebrity name and their handle, to try and capture as much as possible without collecting noise). I summed the number of tweets from the second to the minute level using aggregate().
The use of #goldenglobes has a hashtag was fairly volatile over the course of the night. I suspect this is because use of the hashtag increases when a winner of an award is announced. A quick glance at the top winners of the night seem to confirm this, as tweets about Lady Gaga, Sandra Oh, and Malek Rami increased immediately after their win was announced.
Curiously, Variety did not mention Bohemian Rhapsody winning the (Drama) Best Picture award as a “top most-discussed” moment, despite producing the largest peak of the night (a more in-depth analysis of that time span suggests a compounding effect—Rami winning the “Best Actor” award coincided with Bohemian Rhapsody winning “Best Motion Picture”, so the two trend somewhat closely.
However, mentions of these major actors and media products are only a blimp of the overall Golden Globes coverage. This is an important consideration, because the “most tweeted about celebrities” were tweeted about specifically because they won awards. Otherwise, mentions of them were simply rumbles in Twitterspace (as indicated by the relatively low volume during non-winning times).
Another interesting feature of these award-celebrating tweets is the long-lag. For example, tweets about Sandra Oh and Lady Gaga continued well after their respective wins (one reason for Sandra Oh’s sustained attention may also be her role as role.
The second, smaller blip for Lady Gaga occurs when she does not win the Best Actress award for Shallow (the award went to Glen Close). This “bliping” is actually also a pattern seen among other surprising “losers” of the night (e.g., Crazy Rich Asians, Black Panther).
To advance my ggplot2 skills (“skillz!”), I decided to use reshape2’s melt() function, which helps transform wide data to long data (you can learn more about it here and here). This was a relatively new tool for me, but learning it gave me a lot more control over the multiple lines in my graph!
This was also my second time using plotly. This was my first time embedding annotations via plotly, which was a really interesting learning experience. For those who are familiar with ggplot2 and are interested in creating interactive data, I strongly encourage learning plotly! I still feel like I’m in the learning stage… but I know that the more I use it, the better I become!