Thanks to the phenomenal data science site KDnuggets, we won free registration as an underrepresented group to attend April’s Southern Data Science Conference (SDSC). We were BEYOND excited. As PhD students studying Botany, we create networks to learn how high-dimensional biological data (microbial, genetic, metabolic, etc.) can be used to predict complex plant traits and symbiotic relationships. While most people at the conference had to ask twice, “Wait, what do you study?!” we were thrilled for the opportunity to learn more about data science techniques, eager to extract meaningful information from the wealth of data we had acquired. We were not disappointed!
The energy at SDSC was electric. Friendships were immediately forged—uniting people across the globe, from Timbuktu to Wyoming. We met data science students and leaders ranging from industry to academia. The single track meeting made it possible to attend every talk and the coffee and food breaks were perfect to create a friendly atmosphere to pose questions, learn, and network.
There were so many enlightening talks at SDSC, but one of our favorites was on emoji analyses using Twitter. Hamdan Azhar of PRISMOJI used emojis as a gauge for sentiment analysis; from pop culture to politics. The talk started as light hearted and a little playful, but ended with extreme insight. We left Atlanta thinking about how language is dynamic and emojis can be a powerful tool for understanding human emotion and cultural perception. In fact, Oxford Dictionary's word of the year was the crying emoji in 2015! (https://en.oxforddictionaries.com/word-of-the-year/word-of-the-year-2015)
Azhar’s talk left such an impression on us that we decided to use his tutorial (https://prismoji.com/2017/02/06/emoji-data-science-in-r-tutorial/) to perform our own emoji analysis.
Packages like twitteR allow for emoji analysis in R by analyzing the emojis in social posts in the same way people can search and filter text. We can filter emoji use based on a trending hashtag, a brand, or even location. We performed our first emoji analysis by exploring the most popular emojis associated with the hashtag `womeninstem`.
We can see here the emojis are science related, female centric, and filled with positive accolades - all emojis we’d associate with our experience at SDSC!
Interested in the code we wrote to produce this post? Check out our code here (https://bayesianbabes.netlify.com/post/i-heart-emoji-statistics/)!