홈 › Papers

All in One: A Multi-Task Learning for Emoji, Sentiment and Emotion Analysis in Code-Mixed Text

Code mixed language and emojis are being extensively used in social media to express opinions. In this paper, we propose a novel task that focuses on suggesting appropriate emojis in English-Hindi code-mixed sentences. We aim to exploit the dependency between emotion, sentiment, and emojis for building an end-to-end framework that can simultaneously identify the emotion, sentiment and emojis in code-mixed sentences. We introduce the Code-Mixed Emoji, Emotion and Sentiment aware Dataset (CMEESD) which is an extension of the SemEval 2020 Task 9. We establish strong baselines to predict the correct emojis by simultaneously identifying the emotion and sentiment of a given tweet. The sentiment and emotion prediction in turn helps for the appropriate emoji classification. Empirical results on the CMEESD dataset demonstrate that the proposed multi-task framework yields better performance over the single-task framework.

Code (0)

등록된 구현이 없습니다.

Tasks

Methods 이 논문이 사용한 방법론

Similar Papers 제목 키워드 기반