Sentiment Polarity Estimation of Emoticons by Polarity Scoring of Character Components

摘要

Emoticons are frequently used in online communication. This study, which focuses on front-facing emoticons (Japanese style), proposes an easy method for estimating the sentiment polarity (positive, negative, or neutral) of emoticons by polarity scoring of characters that correspond with features of the emoticon (cheeks, eyes, or mouth). The polarity score is calculated based on the appearance probability of emoticon characters.