Emojis might be considered the digital extension of facial expressions, a universal form of communication that even infants can understand. However, while the idea of emojis unifying people across language barriers is enticing, people of different cultures might not use emojis in the same way.
In a recent study, Penn Engineers used machine learning to investigate similarities and differences in emoji usage in countries they defined as having either “Eastern” or “Western” cultures and what those differences might mean for how scientists view the relationship between online communication and health.
For example, when discussing health-related topics, the Western cultures tended to use more emojis with negative facial expressions than the Eastern cultures, which tended to use emojis representing health care-related objects, like a pill or hypodermic needle. An awareness that certain groups may put more emphasis on emotions and others on treatment logistics could be useful information for doctors or organizations that run health campaigns.
Lyle Ungar, a professor in the Department of Computer and Information Science, and Sharath Chandra Guntuku, previously a postdoctoral researcher at Penn Medicine and Penn Engineering and currently a research scientist at Penn Medicine’s Center for Digital Health, led the research with Mingyang Li, then a graduate student in Penn Engineering’s MSE in Data Science program.
Ungar and his colleagues’ research aims to understand how people’s online language, images, and behavior relate to their mental and physical health.
This story was originally written for Penn Engineering by Lauren Salig.