Abstract
This study analyses a large social media corpus to explore the use of emojis by men and women. Starting with a large raw corpus of Facebook comments, we filtered that to 60,993 comments which include emojis and have been labelled by author’s sex. The objective of the paper is twofold: (a) to explore if men and women use emojis differently in online communication; and (b) to explore if we can predict authorship and emojis most likely to be used by men or women. We applied various analytical measures and machine learning techniques. Findings reveal that females use emojis more frequently and consistently, favouring emotional and expressive emojis. Males use fewer emojis but exhibit higher emoji use per comment, often selecting positive and affirmative symbols. Machine learning successfully distinguishes authorship of emoji usage, and the odds ratio identifies emojis strongly associated with males and females, aligning with social expectations and stereotypes.