本研究主要透過共現圖理解語句中同時出現的表情符號所包含的情緒，因表情符號的應用在其他自然語言處理的任務上非常重要。在不同的社會族群中，表情符號的解釋可能都有所不同。由於單一表情符號可以有不同的解釋方式，因此本篇研究認為單一表情符號難以被簡單分類，而是能代表不只一種的情緒和意義。

我們考慮到不同語言的資料，特別是針對語料資源匱乏的語言。此研究不使用文本內容作為訓練資料，而是提取了推文中使用的表情符號，及參考Ekmans的基本情緒分類為情緒類別依據。

本研究使用無監督學習的方法，將表情符號以建立共現圖的方式，映射到上述提到的情緒類別中。通過觀察推文中表情符號如何表達情感，我們可以了解某些社會族群的人們如何使用表情符號交流。

It is important to understand the emotions that are contained within emojis before we can use the emojis for NLP task, since the interpretation of emojis can be different in each socio-demographics group. Since emojis can be interpreted differently, this work does not see that emojis has to be classified rigidly, instead it can have fuzzy level of emotions. With consideration of how it can be used in any language, especially in low-resource languages, this work extracted the emotion tendencies that are contained in the emojis used in tweets without using text as training data and referred to Ekmans's basic emotions as emotion components. This work used unsupervised approach for mapping of emojis into basic emotions. The mapping process involves making graph from the co-occurrence of emoji in tweets. By observing how emotions are represented by emojis in tweets, we can gain better understanding how the people in certain demographics generally use the emojis in text communication.

1. Manish Reddy Yedulla Anurag Illendula. Learning emoji embeddings using emoji co-occurrence network graph. International Workshop on Emoji Understanding and Applications in Social Media, 2018.
2. Yuita Arum Sari, Evy Ratnasari, Siti Mutrofin, and Agus Zainal Arifin. User emotion identification in twitter using specific features: Hashtag, emoji, emoticon, and adjective term. Jurnal Ilmu Komputer dan Informasi, 7:18–23, 08 2014.
3. Anders Sgaard Iyad Rahwan Sune Lehmann Bjarke Felbo, Alan Mislove. Using millions of emoji occurrences to learn any-domain representations for detecting sentiment, emotion and sarcasm. Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, 2017.
4. Dmitry Davidov, Oren Tsur, and Ari Rappoport. Enhanced sentiment learning using twitter hashtags and smileys. In Proceedings of the 23rd international conference on computational linguistics: posters, pages 241–249. Association for Computational Linguistics, 2010.
5. Ben Eisner, Tim Rockt¨aschel, Isabelle Augenstein, Matko Bosnjak, and Sebastian Riedel. emoji2vec: Learning emoji representations from their description. In Proceedings of The Fourth International Workshop on Natural Language Processing for Social Media, pages 48–54, Austin, TX, USA, November 2016. Association for Computational
Linguistics.
6. Francesco Ronzano Horacio Saggion Francesco Barbieri, German Kruszewski. How cosmopolitan are emojis?: Exploring emojis usage and meaning over different languages with distributional semantics. MM '16: ACM Multimedia Conference, pages 531–535, 2016.
7. Alec Go, Richa Bhayani, and Lei Huang. Twitter sentiment classification using distant supervision. CS224N Project Report, Stanford, 1(2009):12, 2009.
8. Ga¨el Guibon, Magalie Ochs, and Patrice Bellot. From Emoji Usage to Categorical Emoji Prediction. In 19th International Conference on Computational Linguistics and Intelligent Text Processing (CICLING 2018), Hano¨ı, Vietnam, March 2018.
9. John Hannon, Mike Bennett, and Barry Smyth. Recommending twitter users to follow using content and collaborative filtering approaches. In In RecSys 2010, pages 26–30. ACM, 2010.
10. Xia Hu, Jiliang Tang, Huiji Gao, and Huan Liu. Unsupervised sentiment analysis with emotional signals. In Proceedings of the 22nd international conference on WorldWide Web, pages 607–618. ACM, 2013.
11. Wegdan A. Hussien, Mahmoud Al-Ayyoub, Yahya M. Tashtoush, and Mohammed Al-Kabi. On the use of emojis to train emotion classifiers. CoRR, abs/1902.08906, 2019.
12. Lisa Lebduska. Emoji, emoji, what for art thou? 2014.
13. Kris Markman and Sae Oshima. Pragmatic play? some possible functions of english
emoticons and japanese kaomoji in computer-mediated discourse. volume 8, 01 2007.
14. Hannah Miller, Daniel Kluver, Jacob Thebault-Spieker, Loren Terveen, and Brent Hecht. Understanding emoji ambiguity in context: The role of text in emoji-related miscommunication. 05 2017.
15. Hannah Miller, Jacob Thebault-Spieker, Shuo Chang, Isaac Johnson, Loren Terveen, and Brent Hecht. ”blissfully happy” or ”ready to fight”: Varying interpretations of emoji. In Proceedings of the 10th International Conference onWeb and Social Media, ICWSM 2016, Proceedings of the 10th International Conference on Web and Social
Media, ICWSM 2016, pages 259–268. AAAI press, 1 2016.
16. Jaram Park, Vladimir Barash, Clay Fink, and Meeyoung Cha. Emoticon style: Interpreting differences in emoticons across cultures. 07 2013.
17. A. K. Rakhmetullina, Dietrich Trautmann, and Georg Groh. Distant supervision for emotion classification task using emoji 2 emotion. 2018.
18. Jonathon Read. Using emoticons to reduce dependency in machine learning techniques
for sentiment classification. In Proceedings of the ACL Student ResearchWorkshop, pages 43–48, Ann Arbor, Michigan, June 2005. Association for Computational Linguistics.
19. Ian Wood and S Ruder. Emoji as emotion tags for tweets. 05 2016.
20. Shaochen Wu. A dynamic influence keyword model for identifying implicit user interests on social networks. Master’s thesis, National Tsing Hua University, Hsinchu, Taiwan, 2013.