---
title: "ArCorona — paper by Mubarak and Hassan (2021)"
description: "The authors address the lack of sizable, hand-labeled Arabic COVID-19 tweet data that could… 60.5 F1. 23 references, cited by 3 papers here."
authors: ["Hamdy Mubarak", "Sabit Hassan"]
year: 2021
venue: "Proceedings of the 12th International Workshop on Health Text Mining and Information Analysis"
source: "https://aclanthology.org/2021.louhi-1.1/"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/mubarak-2021-arcorona-analyzing-arabic-tweets-early/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/mubarak-2021-arcorona-analyzing-arabic-tweets-early/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2021](https://md.research.elamur.ai/free/en/papers/2021/page/7/) / ArCorona: Analyzing Arabic Tweets in the Early Days of Coronavirus (COVID-19) Pandemic

# ArCorona: Analyzing Arabic Tweets in the Early Days of Coronavirus (COVID-19) Pandemic

**Authors:** [Hamdy Mubarak](https://md.research.elamur.ai/free/en/authors/hamdy-mubarak/) and [Sabit Hassan](https://md.research.elamur.ai/free/en/authors/sabit-hassan/) · **Year:** 2021 · **Venue:** [Proceedings of the 12th International Workshop on Health Text Mining and Information Analysis](https://md.research.elamur.ai/free/en/venues/international-workshop-on-health-text-mining-and-information-analysis-louhi/) ([International Workshop on Health Text Mining and Information Analysis 2021](https://md.research.elamur.ai/free/en/venues/international-workshop-on-health-text-mining-and-information-analysis-louhi/2021/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2021.louhi-1.1/) · [PDF](https://aclanthology.org/2021.louhi-1.1.pdf) · [OpenAlex](https://openalex.org/W3108878757)

**Citations:** 7 citations (OpenAlex, 2026-10-14) · [3](https://md.research.elamur.ai/free/en/papers/mubarak-2021-arcorona-analyzing-arabic-tweets-early/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
The authors address the lack of sizable, hand-labeled Arabic COVID-19 tweet data that could help governments track public concerns, requests, and harmful content such as rumors and bogus cures. They gathered about 30M tweets from February–March 2020 by searching for the colloquial Arabic word for "Corona," which ordinary users favor over the official term. For each of 40 days, they kept the 200 most-retweeted tweets, giving 8,000 items. A native speaker labeled these into 13 classes, including reports, government actions, rumors, advice, cures, and prayers. The classes are also grouped into a coarser important vs. LessImportant split. The paper analyzes class trends over time, topics, hate speech, and the countries tweets came from. It then compares SVMs on n-gram and embedding features against fine-tuned AraBERT. A character n-gram SVM performs about as well as AraBERT on the binary task, while AraBERT is clearly better on the 13-way task. The released ArCorona set gives Arabic crisis informatics a reusable, manually checked benchmark.

## Key points
- **ArCorona dataset:** 8,000 tweets (Feb 21–Mar 31, 2020), 13 classes, released as tweet id, date and label. Cohen's kappa on 200 double-annotated tweets was 0.85.
- **Class groups:** the first 8 classes (REP, ACT, INFO, RUMOR, ADVICE, SEEK ACT, CURE, VOLUNT) count as important. PRSNL, SUPPORT, PRAYER, UNIMP and NOT ARB merge into LessImportant.
- **Multi-class tweets:** roughly 10% of tweets fit several classes. These receive the first applicable important class.
- **Data composition:** Saudi Arabia accounts for about 59% of original authors. A 1,000-tweet topic sample found about 7% hate speech, for example blaming China and Iran.
- **Classification results:** with a 6000/1000/1000 split, the binary task reached a best macro-F1 of 79.8 (SVM, char [2-5] tf-idf), with AraBERT at 79.6. In the fine-grained task, AraBERT led with 60.5 F1, against 56.3 for the word [1-2] SVM.

## Interesting ideas
- Picking the top-retweeted items for each day, rather than sampling the whole pool, yields an annotation set that is balanced over time and focused on content people actually saw.
- When an item fits several labels, a fixed priority order of classes picks one label deterministically. This keeps single-label data consistent while openly noting the multi-label cases as a limitation.
- Annotators were told to open attached images, videos, or links and label by the whole message, since the attachment can override what the text alone suggests.
- Defining fine classes that collapse into a coarse binary split lets one dataset be scored at two granularities. This shows where cheap models suffice and where richer ones pay off.
- A majority-class baseline and a character n-gram SVM were kept beside the transformer, and the simple model nearly matched it on the easier task.

## For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

- **I1** (learn): Build silver or evaluation samples stratified by source and period, so that frequent but unrepresentative text does not dominate training or testing.
- **I2** (frames): When an utterance carries several speech acts, an explicit, logged priority rule can choose the primary act and record the others, keeping the decision explainable.
- **I3** (world): Material from quoted or linked context should be able to revise the inferred state of the world, rather than relying on the surface sentence alone.
- **I4** (learn): Report results for both coarse and fine role or frame categories, so that gains from heavier parsers are measured exactly where they matter.
- **I5** (learn): Always run trivial and shallow baselines as negative controls, to show that a structured, explainable pipeline beats them for real rather than by chance.

*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

- [Hamad bin Khalifa University](https://md.research.elamur.ai/free/en/universities/hamad-bin-khalifa-university/) — University, Doha, Qatar · 141 works · 14,175 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🤝 Social Sciences](https://md.research.elamur.ai/free/en/shelves/social-sciences/) · 1,772 works · [🤝 Sociology and Political Science](https://md.research.elamur.ai/free/en/shelves/social-sciences/sociology-and-political-science/) · 577 works

🧭 Topics (1): [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 2,908 works

## Cite

Cite as: Hamdy Mubarak and Sabit Hassan. 2021. ArCorona: Analyzing Arabic Tweets in the Early Days of Coronavirus (COVID-19) Pandemic. Proceedings of the 12th International Workshop on Health Text Mining and Information Analysis.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{mubarak2021arcorona,
  title = {{ArCorona: Analyzing Arabic Tweets in the Early Days of Coronavirus (COVID-19) Pandemic}},
  author = {Hamdy Mubarak and Sabit Hassan},
  booktitle = {Proceedings of the 12th International Workshop on Health Text Mining and Information Analysis},
  year = {2021},
  url = {https://aclanthology.org/2021.louhi-1.1/}
}
```

- [← previous Anaphora Resolution in Dialogue: Description of the DFKI-TalkingRobots System for the CODI-CRAC 2021 Shared-Task](https://md.research.elamur.ai/free/en/papers/anikina-2021-anaphora-resolution-dialogue-description-dfki/)
- [next → Cascading Adaptors to Leverage English Data to Improve Performance of Question Answering for Low-Resource Languages](https://md.research.elamur.ai/free/en/papers/pandya-2021-cascading-adaptors-leverage-english-data/)
- [↑ back to papers of 2021](https://md.research.elamur.ai/free/en/papers/2021/page/7/)

## References

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

1. Ibrahim Abu Farha and Walid Magdy. 2019. [Mazajak: An online Arabic sentiment analyser](https://md.research.elamur.ai/free/en/papers/farha-2019-mazajak-online-arabic-sentiment-analyser/). In Proceedings of the Fourth Arabic Natural Language Processing Workshop, pages 192–198, Florence, Italy. Association for Computational Linguistics.
2. Fine-grained Classification Firoj Alam, Shaden Shaar, Fahim Dalvi, Hassan Sajjad, Alex Nikolov, Hamdy Mubarak, Giovanni Da San Martino, Ahmed Abdelali, Nadir Durrani, Kareem Darwish, and Preslav Nakov. 2020. Fighting the covid-19 infodemic: Modeling the perspective of journalists, fact-checkers, social media platforms, policy makers, and the society. arXiv:2005.00033. Our next set of experiments were designed for fine-grained classification for 13 classes. With F1 score of 60.5, AraBERT outperformed others (Table 4). Error Analysis: AraBERT confusion matrix (Figure 6) shows that PRSNL, INFO and RUMOR are the hardest classes to identify and the most common error is misclassifying INFO as ADVICE. We hypothesize these errors can be reduced if larger data set is being annotated. [↗](https://arxiv.org/abs/2005.00033)
3. Sarah Alqurashi, Ahmad Alhindi, and Eisa Alanazi. 2020. [Large arabic twitter dataset on covid-19](https://md.research.elamur.ai/free/en/papers/alqurashi-2020-large-arabic-twitter-dataset-covid/). arXiv preprint arXiv:2004.04315. [↗](https://arxiv.org/abs/2004.04315)
4. Thayer Alshaabi, David R. Dewhurst, Joshua R. Minot, Michael V. Arnold, Jane L. Adams, Christopher M. Danforth, and Peter Sheridan Dodds. 2020. The growing amplification of social media: Measuring temporal and social contagion dynamics for over 150 languages on twitter for 2009–2020. Available online at http://arxiv.org/abs/2003.03667. Conclusion and Future Work We present the largest publicly available manually annotated dataset of Arabic tweets for 13 classes that includes the most retweeted tweets in the early Lama Alsudias and Paul Rayson. 2020. [COVID-19 and Arabic Twitter: How can Arab world governments and public health organizations learn from social media?](https://md.research.elamur.ai/free/en/papers/alsudias-2020-covid-19-arabic-twitter-how/) In Proceedings of the 1st Workshop on NLP https://huggingface.co/for COVID-19 at ACL 2020, Online. Association for Computational Linguistics. Marseille, France. European Language Resource Association. [↗](https://arxiv.org/abs/2003.03667)
5. Wissam Antoun, Fady Baly, and Hazem M. Hajj. 2020. [Arabert: Transformer-based model for arabic language understanding](https://md.research.elamur.ai/free/en/papers/antoun-2020-arabert-transformer-based-model-arabic/). ArXiv, abs/2003.00104. [↗](https://arxiv.org/abs/2003.00104)
6. Binxuan Huang and Kathleen M Carley. 2020. Disinformation and misinformation on twitter during the novel coronavirus outbreak. arXiv preprint arXiv:2006.04278. [↗](https://arxiv.org/abs/2006.04278)
7. Juan M Banda, Ramya Tekumalla, Guanyu Wang, Jingyuan Yu, Tuo Liu, Yuning Ding, and Gerardo Chowell. 2020. A large-scale covid-19 twitter chatter dataset for open scientific research– an international collaboration. arXiv preprint arXiv:2004.03688. [↗](https://arxiv.org/abs/2004.03688)
8. Yan Leng, Yujia Zhai, Shaojing Sun, Yifei Wu, Jordan Selzer, Sharon Strover, Julia Fensel, Alex Pentland, and Ying Ding. 2020. Analysis of misinformation during the covid-19 outbreak in china: cultural, social and political entanglements. arXiv preprint arXiv:2005.10414. [↗](https://arxiv.org/abs/2005.10414)
9. Houda Bouamor, Sabit Hassan, and Nizar Habash. 2019. [The MADAR shared task on Arabic fine-grained dialect identification](https://md.research.elamur.ai/free/en/papers/bouamor-2019-madar-shared-task-arabic-fine/). In Proceedings of the Fourth Arabic Natural Language Processing Workshop, pages 199–207, Florence, Italy. Association for Computational Linguistics.
10. L. Li, Q. Zhang, X. Wang, J. Zhang, T. Wang, T. Gao, W. Duan, K. K. Tsoi, and F. Wang. 2020. Characterizing the propagation of situational information in social media during covid-19 epidemic: A case study on weibo. IEEE Transactions on Computational Social Systems, 7(2):556–562. [↗](https://doi.org/10.1109/tcss.2020.2980007)
11. Emily Chen, Kristina Lerman, and Emilio Ferrara. 2020. Tracking social media discourse about the covid-19 pandemic: Development of a public coronavirus twitter data set. JMIR Public Health Surveill, 6(2):e19273. [↗](https://doi.org/10.2196/19273)
12. Arun S Maiya. 2020. ktrain: A low-code library for augmented machine learning. arXiv preprint arXiv:2004.10703. [↗](https://arxiv.org/abs/2004.10703)
13. Hamdy Mubarak, Ahmed Abdelali, Sabit Hassan, and Kareem Darwish. 2020. Spam detection on arabic twitter. In Social Informatics, pages 237–251, Cham. Springer International Publishing. [find](https://scholar.google.com/scholar?q=Hamdy+Mubarak%2C+Ahmed+Abdelali%2C+Sabit+Hassan%2C+and+Kareem+Darwish%2E+2020%2E+Spam+detection+on+arabic+twitter%2E+In+Social+Informatics%2C+pages+237%E2%80%93251%2C+Cham%2E+Springer+International+Publishing%2E)
14. Matteo Cinelli, Walter Quattrociocchi, Alessandro Galeazzi, Carlo Michele Valensise, Emanuele Brugnoli, Ana Lucia Schmidt, Paola Zola, Fabiana Zollo, and Antonio Scala. 2020. The covid-19 social media infodemic. arXiv preprint arXiv:2003.05004. [↗](https://arxiv.org/abs/2003.05004)
15. Hamdy Mubarak, Sabit Hassan, and Ahmed Abdelali. 2021. Adult content detection on arabic twitter: Analysis and experiments. In Proceedings of the Sixth Arabic Natural Language Processing Workshop. [find](https://scholar.google.com/scholar?q=Hamdy+Mubarak%2C+Sabit+Hassan%2C+and+Ahmed+Abdelali%2E+2021%2E+Adult+content+detection+on+arabic+twitter%3A+Analysis+and+experiments%2E+In+Proceedings+of+the+Sixth+Arabic+Natural+Language+Processing+Workshop%2E)
16. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. [BERT: Pre-training of deep bidirectional transformers for language understanding](https://md.research.elamur.ai/free/en/papers/devlin-2019-bert-pre-training-deep-bidirectional/). In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), pages 4171–4186, Minneapolis, Minnesota. Association for Computational Linguistics. [↗](https://doi.org/10.18653/v1/n19-1423)
17. Umair Qazi, Muhammad Imran, and Ferda Ofli. 2020. Geocov19. SIGSPATIAL Special, 12(1):6–15. [find](https://scholar.google.com/scholar?q=Umair+Qazi%2C+Muhammad+Imran%2C+and+Ferda+Ofli%2E+2020%2E+Geocov19%2E+SIGSPATIAL+Special%2C+12%281%29%3A6%E2%80%9315%2E)
18. Gautam Kishore Shahi, Anne Dirkson, and Tim A Majchrzak. 2020. An exploratory study of covid-19 misinformation on twitter. arXiv preprint arXiv:2005.05710. [↗](https://arxiv.org/abs/2005.05710)
19. Fatima Haouari, Maram Hasanain, Reem Suwaileh, and Tamer Elsayed. 2020. [ArCOV-19: The first Arabic COVID-19 twitter dataset with propagation networks](https://md.research.elamur.ai/free/en/papers/haouari-2021-arcov-19-first-arabic-covid/). arXiv:2004.05861. [↗](https://arxiv.org/abs/2004.05861)
20. Qiang Yang, Hind Alamro, Somayah Albaradei, Adil Salhi, Xiaoting Lv, Changsheng Ma, Manal Alshehri, Inji Jaber, Faroug Tifratene, Wei Wang, et al. 2020. Senwave: Monitoring the global sentiments under the covid-19 pandemic. arXiv preprint arXiv:2006.10842. [↗](https://arxiv.org/abs/2006.10842)
21. Sabit Hassan, Hamdy Mubarak, Ahmed Abdelali, and Kareem Darwish. 2021. [Asad: Arabic social media analytics and understanding](https://md.research.elamur.ai/free/en/papers/hassan-2021-asad-arabic-social-media-analytics/). In Proceedings of the Software Demonstrations of the 16th Conference of the European Chapter of the Association for Computational Linguistics.
22. Sabit Hassan, Younes Samih, Hamdy Mubarak, and Ahmed Abdelali. 2020a. [ALT at SemEval-2020 task 12: Arabic and English offensive language identification in social media](https://md.research.elamur.ai/free/en/papers/hassan-2020-alt-semeval-2020-task-12/). In Proceedings of the Fourteenth Workshop on Semantic Evaluation, pages 1891–1897, Barcelona (online). International Committee for Computational Linguistics.
23. Sabit Hassan, Younes Samih, Hamdy Mubarak, Ahmed Abdelali, Ammar Rashed, and Shammur Absar Chowdhury. 2020b. [ALT submission for OSACT shared task on offensive language detection](https://md.research.elamur.ai/free/en/papers/hassan-2020-alt-submission-osact-shared-task/). In Proceedings of the 4th Workshop on Open-Source Arabic Corpora and Processing Tools, with a Shared Task on Offensive Language Detection, pages 61–65,

*Updated 1 November 2026*