In 2015, the Oxford English Dictionary named 😂, aka the “Face with Tears of Joy” emoji, as its word of the year. This news may have prompted you to feel like 🤣, 🤔, 🙄, 😒, 😠 or 😃. Emoticons and emojis are widely attributed as emblems of Internet communication, but, contrary to popular belief, emoticons are actually not a new invention. At the end of the nineteenth century, Ambrose Bierce, an American editorialist, suggested different ways to manipulate punctuation to better represent tone. He proposed using an open parenthesis flipped on its side to express wry smiles and remarks. Throughout the next century, many writers hinted at the benefit of having a special set of symbols to indicate emotion, tone or intention, but it remained a fringe movement for over a century. Then, with the advent of the Internet, the emoticon went viral.
Although emoticons and emojis found their first home on social networking platforms, they weren’t homebodies for long. As speakers embraced their utility, these new punctuation marks quickly infiltrated other communicative spaces on both digital and analog platforms. You’ll spot emojis in Amazon reviews, survey responses, private chat dialogues and even handwritten notes. At first, they were substitutes for other concepts; they now have taken on meanings of their own.
Emojis go far beyond faces, though. In 2016, the avocado emoji gave us a new way to communicate our brunch plans. In 2018, the badger, lobster and llama emojis gave us new ways to describe our trips to the zoo or the aquarium. But, there’s more than meets the 👁️. Many of these emojis have multiple meanings that add unique dimensions to analysis. In the summer of 2018, we sought to analyze social comments about the FIFA World Cup when suddenly we found ourselves confronted by a fierce herd of goats: 🐐. Our brief confusion subsided as we realized the use of the emoji revealed which player each country believed was deserving of the Greatest Of All Time (G.O.A.T.) moniker! Similarly, you might look for 🔥to determine which products socially hip customers find attractive or exemplary, as in “lit,” or 💯 where they are in full agreement with a new idea or campaign. Keep your eyes peeled for innovative uses of emojis; you never know what you’ll find!
In Clarabridge’s NLP engine, emojis and emoticons are treated the same as words. Variations on common emojis or emoticons are mapped to standard tokens. For example, 🙂 and 😀 both map to :). These tokens can be tuned to contain sentiment. The meaning of most emoticons and emojis are context-sensitive. However, there is a small subset of unambiguous emojis and emoticons to which Clarabridge assigns out-of-the-box positive or negative sentiment values.
Analysts can choose to analyze emojis in much the same way that they would approach words. Here are the top use cases for emoticons/emojis:
1. Emoji Cloud
Create a word cloud of just emojis to explore how customers are reacting to your brand. Emotional faces and other symbols can quickly give you a sense of brand perception and may even clue you in to product uses.
2. Correlation with Products and Services
Drill down on specific products, services and experiences to see which emojis are used in conjunction with each one. Or, drill down on a specific emoji to see which products, services and experiences are being discussed. Understanding reactions to these offerings can assist in designing, marketing and promoting your products and services.
3. Emoji Trends
Follow specific emojis over time to see how emotions and impressions are changing. Take advantage of a viral trend or get ahead of a major customer issue. Emojis can expose feelings that sometimes are lost in words.
Emojis have become ubiquitous in our digital world. With increased adoption and pervasiveness of emojis in all channels, it’s paramount to consider them as full-fledged components of customer feedback. They reveal unique messages and definitely invite a certain degree of playfulness in interpretation of feedback!
Read further about emojis and learn more about how the Clarabridge NLP recognizes other named entities such as products, brands, people, and events in our Debunking Natural Language Processing eBook.