Elamur Research⌕
☰
home / papers / 2026 / From Adoption to Adaptation: Tracing the Diffusion of New Emojis on Twitter

From Adoption to Adaptation: Tracing the Diffusion of New Emojis on Twitter

Paper · 2026 · Yuhang Zhou, Xuan Lu and Wei Ai

Summary

Every Unicode release adds emojis that existing NLP models have never seen. Because these emojis have exact release dates and a fixed form, they make a clean test case for how new language units spread and change meaning. The authors studied English tweets collected through the Twitter API from May 2018 to May 2022, tracking emojis from Unicode 11.0–13.0. They measured how often each emoji was used and estimated tweet sentiment with Vader. Drawing on Rogers' diffusion of innovation theory, they linked later popularity to the size of the early-adopter community and to the emoji's semantics. They also built a language-model-based framework that finds words and older emojis used in similar contexts, which helps interpret each new emoji. In sentiment classification and irony detection, swapping unfamiliar emojis for these familiar surrogates generally improved accuracy, with no retraining needed. This offers a low-cost fix for models facing new emojis and a reusable method for studying emerging non-verbal tokens such as memes or hashtags.

Key points

  • Emoji popularity is highly skewed, close to a power law. Holidays such as New Year's Day and Valentine's Day cause short usage spikes, and older versions can level off or decline as newer emojis arrive.
  • Pleading face changed meaning over time. Its average Vader score moved from −0.198 in October 2018 to 0.221 in October 2019, and the words most associated with it by PMI shifted from "sad" and "hate" to "prettiest" and "cutest."
  • Early hashtag popularity, used as a proxy for early-adopter community size, correlated positively with emoji counts 14 months later (Spearman ρ = 0.580, p = 0.08 for Emoji 11.0; ρ = 0.530, p = 0.11 for 12.0). The authors acknowledge that the sample is small and these results are not significant.
  • For a model pretrained before August 2019, replacing new emojis with similar older ones raised sentiment accuracy by 5.02% (sentimental) and 4.71% (entity) for Emoji 13.0, and by 1.78% and 1.51% for Emoji 14.0. These figures are relative improvements.
  • Emoji surrogates beat substitution with emoji names or ChatGPT descriptions in 11 of 16 sentiment settings and 13 of 16 irony settings. Further tests used twitter-roberta-base-2022-154m, whose pretraining data already includes the new emojis.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓University of Maryland, College ParkCollege Park · United States140 works · 29,011 citations🎓University of ArizonaTucson · United States39 works · 10,603 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksHuman-Computer Interaction · 160 works

Cite

Cite as Yuhang Zhou, Xuan Lu and Wei Ai. 2026. From Adoption to Adaptation: Tracing the Diffusion of New Emojis on Twitter. Proceedings of the Seventh Workshop on Natural Language Processing and Computational Social Science.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.nlpcss-1.17

@inproceedings{zhou2026adoption,
  title = {{From Adoption to Adaptation: Tracing the Diffusion of New Emojis on Twitter}},
  author = {Yuhang Zhou, Xuan Lu and Wei Ai},
  booktitle = {Proceedings of the Seventh Workshop on Natural Language Processing and Computational Social Science},
  year = {2026},
  doi = {10.18653/v1/2026.nlpcss-1.17},
  url = {https://doi.org/10.18653/v1/2026.nlpcss-1.17}
}

References

45 references, 10 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Oshin Agarwal and Ani Nenkova. 2022. Temporal effects on pre-trained models for language processing tasks. Transactions of the Association for Computational Linguistics, 10:904–921.
  2. Wei Ai, Xuan Lu, Xuanzhe Liu, Ning Wang, Gang Huang, and Qiaozhu Mei. 2017. Untangling emoji popularity through semantic embeddings. In ICWSM 2017. find
  3. Francesco Barbieri, Jose Camacho-Collados, Leonardo Neves, and Luis Espinosa-Anke. 2020. Tweeteval: Unified benchmark and comparative evaluation for tweet classification. arXiv preprint arXiv:2010.12421.
  4. Zhenpeng Chen, Sheng Shen, Ziniu Hu, Xuan Lu, Qiaozhu Mei, and Xuanzhe Liu. 2019. Emoji-Powered representation learning for Cross-Lingual sentiment classification. In WWW 2019. find
  5. Kevin Clark, Urvashi Khandelwal, Omer Levy, and Christopher D Manning. 2019. What does bert look at? an analysis of bert’s attention. arXiv preprint arXiv:1906.04341. find
  6. Israel Cohen, Yiteng Huang, Jingdong Chen, Jacob Benesty, Jacob Benesty, Jingdong Chen, Yiteng Huang, and Israel Cohen. 2009. Pearson correlation coefficient. Noise reduction in speech processing, pages 1–4. find
  7. Riley Crane and Didier Sornette. 2008. Robust dynamic classes revealed by measuring the response function of a social system. Proceedings of the National Academy of Sciences, 105(41):15649–15653. find
  8. Ben Eisner, Tim Rocktäschel, Isabelle Augenstein, Matko Bošnjak, and Sebastian Riedel. 2016. emoji2vec: Learning emoji representations from their description. arXiv preprint.
  9. Bjarke Felbo, Alan Mislove, Anders Søgaard, Iyad Rahwan, and Sune Lehmann. 2017. Using millions of emoji occurrences to learn any-domain representations for detecting sentiment, emotion and sarcasm. EMNLP.
  10. Yunhe Feng, Zheng Lu, Wenjun Zhou, Zhibo Wang, and Qing Cao. 2020. New emoji requests from twitter users: when, where, why, and what we can do about them. ACM Transactions on Social Computing, 3(2):1–25. find
  11. Jing Ge. 2019. Emoji sequence use in enacting personal identity. In Companion Proceedings of The 2019 World Wide Web Conference, WWW ’19, page 426–438, New York, NY, USA. Association for Computing Machinery. find
  12. Fabrizio Gilardi, Meysam Alizadeh, and Maël Kubli. 2023. Chatgpt outperforms crowd-workers for text-annotation tasks. arXiv preprint arXiv:2303.15056.
  13. Jack Grieve, Andrea Nini, and Diansheng Guo. 2018. Mapping lexical innovation on american social media. Journal of English Linguistics, 46(4):293–319. find
  14. Colin ME Halverson, Claire E Donnelly, Michael Weiner, and Joy L Lee. 2023. Content analysis of emoji and emoticon use in clinical texting systems. JAMA Network Open, 6(6):e2318140–e2318140. find
  15. Shirley Anugrah Hayati, Aditi Chaudhary, Naoki Otani, and Alan W Black. 2019. What a sunny day: Toward emoji-sensitive irony detection. In Proceedings of the 5th Workshop on Noisy User-generated Text (W-NUT 2019), pages 212–216.
  16. Tianran Hu, Han Guo, Hao Sun, Thuy-vy Nguyen, and Jiebo Luo. 2017. Spice up your chat: the intentions and sentiment effects of using emojis. In ICWSM 2017. find
  17. Xiaolei Huang and Michael Paul. 2018. Examining temporality in document classification. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 694–699. find
  18. Clayton Hutto and Eric Gilbert. 2014. Vader: A parsimonious rule-based model for sentiment analysis of social media text. In ICWSM 2014. find
  19. Michael Johann and Lars Bülow. 2019. One does not simply create a meme: Conditions for the diffusion of internet memes. International Journal of Communication, 13:23. find
  20. Zixuan Ke and Bing Liu. 2022. Continual learning of natural language processing tasks: A survey. arXiv preprint arXiv:2211.12701. find
  21. Mayank Kejriwal, Qile Wang, Hongyu Li, and Lu Wang. 2021. An empirical study of emoji usage on twitter in linguistic and national contexts. Online Social Networks and Media, 24:100149. find
  22. Daniel James Kershaw. 2018. Language change and evolution in online social networks. Lancaster University (United Kingdom). find
  23. Angelika Romanou, Syrielle Montariol, Debjit Paul, Leo Laugier, Karl Aberer, and Antoine Bosselut. 2023. Crab: Assessing the strength of causal relationships between real-world events. arXiv preprint arXiv:2311.04284. find
  24. Shiyue Rong, Weisheng Wang, Umme Ayda Mannan, Eduardo Santana de Almeida, Shurui Zhou, and Iftekhar Ahmed. 2022. An empirical study of emoji use in software development communication. Information and Software Technology. find
  25. Anna Sophie Kümpel, Veronika Karnowski, and Till Keyling. 2015. News sharing in social media: A review of current research on news sharing users, content, and networks. Social media+ society, 1(2):2056305115610141. find
  26. Sara Rosenthal, Preslav Nakov, Svetlana Kiritchenko, Saif Mohammad, Alan Ritter, and Veselin Stoyanov. 2015. Semeval-2015 task 10: Sentiment analysis in twitter. In Proceedings of the 9th international workshop on semantic evaluation (SemEval 2015), pages 451–463.
  27. Victoria Lin, Louis-Philippe Morency, and Eli Ben-Michael. 2023. Text-transport: Toward learning causal effects of natural language. arXiv preprint arXiv:2310.20697. find
  28. Paul Röttger and Janet B Pierrehumbert. 2021. Temporal adaptation of bert and performance on downstream document classification: Insights from social media. arXiv preprint arXiv:2104.08116. find
  29. Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Roberta: A robustly optimized bert pretraining approach. arXiv preprint.
  30. Erin M Steffes and Lawrence E Burgee. 2009. Social ties and online word of mouth. Internet research, 19(1):42–59. find
  31. Daniel Loureiro, Kiamehr Rezaee, Talayeh Riahi, Francesco Barbieri, Leonardo Neves, Luis Espinosa Anke, and Jose Camacho-Collados. 2023. Tweet insights: A visualization platform to extract temporal insights from twitter. arXiv preprint arXiv:2308.02142. — [arXiv:2308.02142](https://arxiv.org/abs/2308.02142) find
  32. Xuan Lu, Wei Ai, Xuanzhe Liu, Qian Li, Ning Wang, Gang Huang, and Qiaozhu Mei. 2016. Learning from the ubiquitous language: An empirical analysis of emoji usage of smartphone users. In UbiComp 2016. find
  33. Xuan Lu, Yanbin Cao, Zhenpeng Chen, and Xuanzhe Liu. 2018. A first look at emoji usage on github: An empirical study. arXiv preprint. find
  34. Zhaochen Su, Zecheng Tang, Xinyan Guan, Juntao Li, Lijun Wu, and Min Zhang. 2022. Improving temporal generalization of pre-trained language models with lexical semantic change. arXiv preprint arXiv:2210.17127. find
  35. Tianlu Wang, Rohit Sridhar, Diyi Yang, and Xuezhi Wang. 2021. Identifying and mitigating spurious correlations for improving robustness in nlp models. arXiv preprint arXiv:2110.07736. find
  36. Lei Yang, Tao Sun, Ming Zhang, and Qiaozhu Mei. 2012. We know what @you \#tag: Does the dual role affect hashtag adoption? In WWW 2012. find
  37. Jerrold H Zar. 2005. Spearman rank correlation. Encyclopedia of biostatistics, 7. find
  38. Long Ma, Chei Sian Lee, and Dion Hoe-Lian Goh. 2014. Understanding news sharing in social media: An explanation from the diffusion of innovations theory. Online information review, 38(5):598–615. find
  39. OpenAI. 2023. Gpt-4 technical report.
  40. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. 2022. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35:27730–27744.
  41. Everett M Rogers, Arvind Singhal, and Margaret M Quinlan. 2014. Diffusion of innovations. In An integrated approach to communication theory and research, pages 432–448. Routledge. find
  42. Fan Zhou, Xovee Xu, Goce Trajcevski, and Kunpeng Zhang. 2021. A survey of information cascade analysis: Models, predictions, and recent advances. ACM Computing Surveys (CSUR), 54(2):1–36. find
  43. Yuhang Zhou and Wei Ai. 2022. \#emoji: A study on the association between emojis and hashtags on twitter. In ICWSM 2022. find
  44. Yuhang Zhou, Paiheng Xu, Xiaoyu Liu, Bang An, Wei Ai, and Furong Huang. 2023. Explore spurious correlations at the concept level in language models for text classification. arXiv preprint arXiv:2311.08648. find
  45. Yuhang Zhou, Paiheng Xu, Xiyao Wang, Xuan Lu, Ge Gao, and Wei Ai. 2024. Emojis decoded: Leveraging chatgpt for enhanced understanding in social media communications. arXiv preprint arXiv:2402.01681. — [arXiv:2402.01681](https://arxiv.org/abs/2402.01681) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zhou-2026-adoption-adaptation-tracing-diffusion-new/