TweetCloud is a Data Visualization project that helps us perform visual analysis on emojis from Twitter to understand public sentiment
Sep 2017 - Nov 2017
Data Visualization team project - Collaboration with Jaina Gandhi
Instructor: Karthik Badam
As a part of our Data Visualization course project, we created a data visualization tool to analyze a public dataset and tell a story. We created a storytelling visualization for emoji usage in Twitter, called Emoji Cloud, and used to extract patterns in emoji associated with popular characters from Game of Thrones
Before coming up with this particular concept, we brainstormed multiple ideas. Since our goal was to tell a story through visualizations, we focused on ideas that solve a problem arising from data that is publicly available.
We finalized on understanding public sentiment from Twitter and more specifically, sentiment arising from emoji usage.
STORYTELLING FROM EMOJI CLOUD
Social media platforms have lately become a popular source for understanding public sentiment due to the freedom of expression they offer. Following real-life events like TV releases and political debates, these platforms create conversations among public, which further form basis for interesting stories of public opinions. Lot of textual analyses on Twitter understand sentiments, networks, user engagement etc. from the tweets. However, emojis are an interesting medium for sentiment expression. Yet, there are not any studies considering emoji visualization in sentiment analysis studies.
According to statistics cited by Ad Week, 92% of the online population uses emojis. Twitter reports that since 2014 alone, over 110 billion emojis have been tweeted.
Data source: Being a fan of Game of Thrones, the popular TV show was my choice of topic for my emoji analysis. We collected emojis associated with characters of Game of Thromes over the last four seasons from Twitter.
Visualization: To tell stories from emojis, we decided to create a visualization tool that captures emojis about popular characters from tweets related to the hashtag #GameOfThrones on Twitter. This tool contains a visualization, Emoji Cloud, which inspired by a word cloud visualizes the unstructured emoji data. The tool gave us lots of interesting stories that reveal public opinions on Game Of Thrones characters.
Clicking a character face triggers the emoji cloud and reshuffles to show the emojis that were used with that character in the tweets. In the emoji cloud visualization, emojis resize based on their occurrences into a concentric layout and bigger emojis for a character inform dominant public sentiments.
Links between faces represent the connection between two characters. Links are interactive and clicking on them restructures emoji cloud to show emojis used with both the characters in a single tweet.
The tool gave us lots of interesting stories that reveal public opinions on Game Of Thrones characters.
1. Arya’s emoji cloud is particularly dominated by emojis of encouragement and strength such as 🙌, 👏, 💪, and 👌. Her dire wolf 🐺 and her sword Needle 🗡 are suggestive of her strong character.
2. A comparison of emoji clouds for Hodor and Joffrey shows how fans can love or hate the fictitious characters in fantasy worlds. All the emojis suggesting grief and heartbreak were used in the tweets relating Hodor’s death whereas emojis suggesting party and celebration were used in tweets pertaining to Joffrey’s death.
Check out my Medium article for more of such insights from our analysis: Tales as told by emojis of Twitter