Abstract

Emojis (e.g., “”) serve as effective tools for expressing emotional states in digital communication. However, empirical evidence on the interpretation of various emojis is scarce in both scientific and practical domains. This study aims to systematically analyze the perceived emotional states of 74 facial emojis by applying the framework of the circumplex model of affect (Russell, 1980) with a Korean sample (N = 1,008). We assessed the perceived valence and arousal levels for each emoji, followed by hierarchical cluster analysis based on these two dimensions. The analysis revealed six distinct clusters characterized by different valence and arousal levels: A strong positive-high arousal sentiment, a moderately positive-medium arousal sentiment, a neutral sentiment with a positive bias, a neutral sentiment with a negative bias, a moderately negative-very low arousal sentiment, and a strong negative-low arousal sentiment. We discuss the patterns of Koreans’ perceived emotional states from emojis and the utility of emojis in global, non-linguistic digital communication, as well as in scientific research investigating everyday emotional experiences.