---
title: "AlexU-AIC at WojoodNER shared task: Sequence Labeling vs"
description: "The paper describes an Alexandria University / Applied Innovation Center entry to the… 91.13% F1. 25 references, cited by 2 papers here."
authors: ["Shereen Elkordi", "Noha Adly", "Marwan Torki"]
year: 2023
venue: "Proceedings of ArabicNLP 2023"
source: "https://aclanthology.org/2023.arabicnlp-1.86/"
doi: "10.18653/v1/2023.arabicnlp-1.86"
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/elkordi-2023-alexu-aic-wojoodner-shared-task/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/elkordi-2023-alexu-aic-wojoodner-shared-task/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2023](https://md.research.elamur.ai/free/en/papers/2023/page/12/) / AlexU-AIC at WojoodNER shared task: Sequence Labeling vs MRC and SWA for Arabic Named Entity Recognition

# AlexU-AIC at WojoodNER shared task: Sequence Labeling vs MRC and SWA for Arabic Named Entity Recognition

**Authors:** Shereen Elkordi, [Noha Adly](https://md.research.elamur.ai/free/en/authors/noha-adly/) and [Marwan Torki](https://md.research.elamur.ai/free/en/authors/marwan-torki/) · **Year:** 2023 · **Venue:** [Proceedings of ArabicNLP 2023](https://md.research.elamur.ai/free/en/venues/arabicnlp/) ([ArabicNLP 2023](https://md.research.elamur.ai/free/en/venues/arabicnlp/2023/)) · **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/2023.arabicnlp-1.86/) · [PDF](https://aclanthology.org/2023.arabicnlp-1.86.pdf) · [DOI 10.18653/v1/2023.arabicnlp-1.86](https://doi.org/10.18653/v1/2023.arabicnlp-1.86) · [OpenAlex](https://openalex.org/W4389518404)

**Citations:** 3 citations (OpenAlex, 2026-10-14) · [2](https://md.research.elamur.ai/free/en/papers/elkordi-2023-alexu-aic-wojoodner-shared-task/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
The paper describes an Alexandria University / Applied Innovation Center entry to the WojoodNER 2023 shared task on Arabic named entity recognition, which had a flat and a nested subtask. The authors compare two framings. The first is classic sequence labeling: a pre-trained Arabic encoder with a token classifier. The second recasts NER as machine reading comprehension (MRC), in the style of Li et al. (2020): each sentence is paired with a query per entity type, and the model extracts answer spans. They test several Arabic encoders and two kinds of queries, translated label keywords versus annotation-guideline descriptions. They also average the weights of the best checkpoints (SWA) instead of building an ensemble. For flat NER, an AraBERT-based MRC model with SWA was their best system, reaching 91.13% F1 and placing eighth. For nested NER, plain sequence labeling with JABER gave 92.61% F1 and third place. The broader finding is that MRC and sequence labeling perform about equally well on Arabic NER. MRC costs much more at inference time, but descriptive queries give it a modest benefit.

## Key points
- **Data:** the Wojood corpus has 21 entity types and 16,817 / 3,133 / 5,990 sentences in train / validation / test. Expanding each sentence into 21 queries grew the MRC splits to 353,157 / 65,793 / 125,790 samples, and most queries have no answer.
- **Class balance:** GPE, Date and Organization each occur more than 11K times. Percentage, Quantity and Unit each occur fewer than 50 times, and Unit is absent from the flat validation set.
- **Encoders:** AraBERTv2, AraBERTv2-Twitter, MARBERT(v2), AraElectra, CamelBERT and JABER were compared. JABER, which uses byte-level BPE, came first and AraBERTv2 second.
- **Flat NER test F1:** sequence labeling with JABER scored 0.9052. MRC with AraBERT scored 0.9065, and adding SWA raised it to 0.9113. Keyword queries did slightly worse than guideline-description queries (about 0.904).
- **Nested NER test F1:** sequence labeling with JABER scored 0.9261, ahead of the best MRC variant at 0.9219 (JABER + SWA). The nested labeler uses 21 parallel B/I/O heads, one per entity type.

## Interesting ideas
- When a span extractor is queried per category, writing each query as the category's guideline definition works better than a bare label name, because the model can draw on the definition's meaning.
- Averaging the weights of a few top checkpoints gave ensemble-like gains at single-model inference cost. A per-class breakdown then showed which checkpoints fixed which classes and which ones degraded rare classes.
- Overlapping per-type predictions were resolved deterministically, with types ranked by training frequency. This rule was then checked against a random tie-break averaged over 5 runs, which showed both scored nearly the same while the fixed rule stayed reproducible.
- The MRC span head pairs candidate starts with candidate ends through an N×N match matrix. This lets it recover several and even overlapping spans for the same query.
- Annotation inconsistencies, such as two equivalent payment services labelled Organisation in one case and Product in the other, were singled out as a main cause of weak classes.

## For Mova

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

- **I1** (frames): role or frame detectors could take the VerbNet/FrameNet definition of a role as their prompt instead of the role label, which would also make each detection easy to explain.
- **I2** (learn): gparse checkpoints could be weight-averaged, and per-relation LAS deltas could be logged to check that rare relations are not silently harmed.
- **I3** (learn): any heuristic conflict-resolution or repair rule could be validated against a random-choice control like this before being kept.
- **I4** (world): entity-mention extraction could use a pairwise start–end matcher to keep nested mentions (an organization inside a location name) as separate world entities.
- **I5** (world): ontology typing from Wikidata/WordNet could flag such near-duplicate entities with conflicting types as candidates for correction in silver data.

*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

- [Alexandria University](https://md.research.elamur.ai/free/en/universities/alexandria-university/) — University, Alexandria, Egypt · 39 works · 287 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 · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works

🧭 Topics (2): [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 2,908 works · [🏷️ Named entity recognition](https://md.research.elamur.ai/free/en/topics/named-entity-recognition/) · 200 works

## Cite

Cite as: Shereen Elkordi, Noha Adly and Marwan Torki. 2023. AlexU-AIC at WojoodNER shared task: Sequence Labeling vs MRC and SWA for Arabic Named Entity Recognition. Proceedings of ArabicNLP 2023.

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

```bibtex
@inproceedings{elkordi2023alexu,
  title = {{AlexU-AIC at WojoodNER shared task: Sequence Labeling vs MRC and SWA for Arabic Named Entity Recognition}},
  author = {Shereen Elkordi, Noha Adly and Marwan Torki},
  booktitle = {Proceedings of ArabicNLP 2023},
  year = {2023},
  doi = {10.18653/v1/2023.arabicnlp-1.86},
  url = {https://doi.org/10.18653/v1/2023.arabicnlp-1.86}
}
```

- [← previous Alex-U 2023 NLP at WojoodNER shared task: AraBINDER (Bi-encoder for Arabic Named Entity Recognition)](https://md.research.elamur.ai/free/en/papers/hussein-2023-alex-u-2023-nlp-wojoodner/)
- [next → Analysis of Corpus-based Word-Order Typological Methods](https://md.research.elamur.ai/free/en/papers/alves-2023-analysis-corpus-based-word-order/)
- [↑ back to papers of 2023](https://md.research.elamur.ai/free/en/papers/2023/page/12/)

## References

25 references, 19 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Muhammad Abdul-Mageed, AbdelRahim Elmadany, and El Moatez Billah Nagoudi. 2021. [ARBERT & MARBERT: Deep bidirectional transformers for Arabic](https://md.research.elamur.ai/free/en/papers/abdul-mageed-2021-arbert-marbert-deep-bidirectional-transformers/). In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 7088–7105, Online. Association for Computational Linguistics. [↗](https://doi.org/10.18653/v1/2021.acl-long.551)
2. Wissam Antoun, Fady Baly, and Hazem 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 preprint arXiv:2003.00104. [↗](https://arxiv.org/abs/2003.00104)
3. Wissam Antoun, Fady Baly, and Hazem Hajj. 2021. [AraELECTRA: Pre-training text discriminators for Arabic language understanding](https://md.research.elamur.ai/free/en/papers/antoun-2021-araelectra-pre-training-text-discriminators/). In Proceedings of the Sixth Arabic Natural Language Processing Workshop, pages 191–195, Kyiv, Ukraine (Virtual). Association for Computational Linguistics. [↗](https://doi.org/10.48550/arxiv.2012.15516)
4. Sarkar Snigdha Sarathi Das, Arzoo Katiyar, Rebecca Passonneau, and Rui Zhang. 2022. [CONTaiNER: Few-shot named entity recognition via contrastive learning](https://md.research.elamur.ai/free/en/papers/das-2022-container-few-shot-named-entity/). In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 6338–6353, Dublin, Ireland. Association for Computational Linguistics.
5. Abbas Ghaddar, Yimeng Wu, Ahmad Rashid, Khalil Bibi, Mehdi Rezagholizadeh, Chao Xing, Yasheng Wang, Duan Xinyu, Zhefeng Wang, Baoxing Huai, et al. 2021. [Jaber and saber: Junior and senior arabic bert](https://md.research.elamur.ai/free/en/papers/ghaddar-2021-jaber-saber-junior-senior-arabic/). arXiv preprint arXiv:2112.04329. [↗](https://arxiv.org/abs/2112.04329)
6. Yucheng Huang, Kai He, Yige Wang, Xianli Zhang, Tieliang Gong, Rui Mao, and Chen Li. 2022. COP-NER: Contrastive learning with prompt guiding for few-shot named entity recognition. In Proceedings of the 29th International Conference on Computational Linguistics, pages 2515–2527, Gyeongju, Republic of Korea. International Committee on Computational Linguistics. [find](https://scholar.google.com/scholar?q=Yucheng+Huang%2C+Kai+He%2C+Yige+Wang%2C+Xianli+Zhang%2C+Tieliang+Gong%2C+Rui+Mao%2C+and+Chen+Li%2E+2022%2E+COP%2DNER%3A+Contrastive+learning+with+prompt+guiding+for+few%2Dshot+named+entity+recognition%2E+In+Proceedings+of+th)
7. Mariam Hussein, Sarah Khaled, Marwan Torki, and Nagwa Elmakky. 2023. [Alex-u 2023 nlp at wojoodner shared task: Arabinder (bi-encoder for arabic named entity recognition)](https://md.research.elamur.ai/free/en/papers/hussein-2023-alex-u-2023-nlp-wojoodner/). In Proceedings of The First Arabic Natural Language Processing Conference (ArabicNLP 2023).
8. Go Inoue, Bashar Alhafni, Nurpeiis Baimukan, Houda Bouamor, and Nizar Habash. 2021. [The interplay of variant, size, and task type in Arabic pre-trained language models](https://md.research.elamur.ai/free/en/papers/inoue-2021-interplay-variant-size-task-type/). In Proceedings of the Sixth Arabic Natural Language Processing Workshop, Kyiv, Ukraine (Online). Association for Computational Linguistics.
9. Pavel Izmailov, Dmitrii Podoprikhin, Timur Garipov, Dmitry Vetrov, and Andrew Gordon Wilson. 2018. [Averaging weights leads to wider optima and better generalization](https://md.research.elamur.ai/free/en/papers/izmailov-2018-averaging-weights-leads-wider-optima/). arXiv preprint arXiv:1803.05407. [↗](https://arxiv.org/abs/1803.05407)
10. Mustafa Jarrar, Muhammad Abdul-Mageed, Mohammed Khalilia, Bashar Talafha, AbdelRahim Elmadany, Nagham Hamad, and Alaa Omer. 2023. [WojoodNER 2023: The First Arabic Named Entity Recognition Shared Task](https://md.research.elamur.ai/free/en/papers/jarrar-2023-wojoodner-2023-first-arabic-named/). In Proceedings of the First Arabic Natural Language Processing Conference (ArabicNLP 2023).
11. Mustafa Jarrar, Mohammed Khalilia, and Sana Ghanem. 2022. [Wojood: Nested arabic named entity corpus and recognition using bert](https://md.research.elamur.ai/free/en/papers/jarrar-2022-wojood-nested-arabic-named-entity/). In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 3626–3636.
12. Arzoo Katiyar and Claire Cardie. 2018. [Nested named entity recognition revisited](https://md.research.elamur.ai/free/en/papers/katiyar-2018-nested-named-entity-recognition-revisited/). In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers), pages 861–871, New Orleans, Louisiana. Association for Computational Linguistics.
13. Jing Li, Aixin Sun, Jianglei Han, and Chenliang Li. 2022. A survey on deep learning for named entity recognition. IEEE Transactions on Knowledge and Data Engineering, 34(1):50–70. [find](https://scholar.google.com/scholar?q=Jing+Li%2C+Aixin+Sun%2C+Jianglei+Han%2C+and+Chenliang+Li%2E+2022%2E+A+survey+on+deep+learning+for+named+entity+recognition%2E+IEEE+Transactions+on+Knowledge+and+Data+Engineering%2C+34%281%29%3A50%E2%80%9370%2E)
14. Xiaoya Li, Jingrong Feng, Yuxian Meng, Qinghong Han, Fei Wu, and Jiwei Li. 2020. [A unified MRC framework for named entity recognition](https://md.research.elamur.ai/free/en/papers/li-2020-unified-mrc-framework-named-entity/). In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 5849–5859, Online. Association for Computational Linguistics.
15. Diego Mollá, Menno Van Zaanen, and Daniel Smith. 2006. Named entity recognition for question answering. In Proceedings of the Australasian language technology workshop 2006, pages 51–58. [find](https://scholar.google.com/scholar?q=Diego+Moll%C3%A1%2C+Menno+Van+Zaanen%2C+and+Daniel+Smith%2E+2006%2E+Named+entity+recognition+for+question+answering%2E+In+Proceedings+of+the+Australasian+language+technology+workshop+2006%2C+pages+51%E2%80%9358%2E)
16. Feng Nan, Ramesh Nallapati, Zhiguo Wang, Cicero Nogueira dos Santos, Henghui Zhu, Dejiao Zhang, Kathleen McKeown, and Bing Xiang. 2021. [Entity-level factual consistency of abstractive text summarization](https://md.research.elamur.ai/free/en/papers/nan-2021-entity-level-factual-consistency-abstractive/). arXiv preprint arXiv:2102.09130. [↗](https://arxiv.org/abs/2102.09130)
17. Artur Nowakowski, Gabriela Pałka, Kamil Guttmann, and Mikołaj Pokrywka. 2022. Adam mickiewicz university at wmt 2022: Ner-assisted and quality-aware neural machine translation. arXiv preprint arXiv:2209.02962. [↗](https://arxiv.org/abs/2209.02962)
18. Xiaoye Qu, Yingjie Gu, Qingrong Xia, Zechang Li, Zhefeng Wang, and Baoxing Huai. 2023. [A survey on arabic named entity recognition: Past, recent advances, and future trends](https://md.research.elamur.ai/free/en/papers/qu-2023-survey-arabic-named-entity-recognition/). arXiv preprint arXiv:2302.03512. [↗](https://arxiv.org/abs/2302.03512)
19. Khaled Shaalan. 2014. [A survey of arabic named entity recognition and classification](https://md.research.elamur.ai/free/en/papers/shaalan-2014-survey-arabic-named-entity-recognition/). Computational Linguistics, 40(2):469–510. [↗](https://doi.org/10.1162/coli_a_00178)
20. Jana Straková, Milan Straka, and Jan Hajic. 2019. [Neural architectures for nested NER through linearization](https://md.research.elamur.ai/free/en/papers/strakova-2019-neural-architectures-nested-ner-through/). In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 5326–5331, Florence, Italy. Association for Computational Linguistics.
21. Hang Yan, Tao Gui, Junqi Dai, Qipeng Guo, Zheng Zhang, and Xipeng Qiu. 2021. [A unified generative framework for various NER subtasks](https://md.research.elamur.ai/free/en/papers/yan-2021-unified-generative-framework-various-ner/). In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 5808–5822, Online. Association for Computational Linguistics. [↗](https://doi.org/10.48550/arxiv.2106.01223)
22. Jie Yang, Shuailong Liang, and Yue Zhang. 2018. [Design challenges and misconceptions in neural sequence labeling](https://md.research.elamur.ai/free/en/papers/yang-2018-design-challenges-misconceptions-neural-sequence/). In Proceedings of the 27th International Conference on Computational Linguistics, pages 3879–3889, Santa Fe, New Mexico, USA. Association for Computational Linguistics. [↗](https://doi.org/10.48550/arxiv.1806.04470)
23. Juntao Yu, Bernd Bohnet, and Massimo Poesio. 2020. [Named entity recognition as dependency parsing](https://md.research.elamur.ai/free/en/papers/yu-2020-named-entity-recognition-as-dependency/). In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 6470– 6476, Online. Association for Computational Linguistics.
24. Sheng Zhang, Hao Cheng, Jianfeng Gao, and Hoifung Poon. 2022. Optimizing bi-encoder for named entity recognition via contrastive learning. arXiv preprint arXiv:2208.14565. [↗](https://arxiv.org/abs/2208.14565)
25. Huiming Zhu, Chunhui He, Yang Fang, and Weidong Xiao. 2020. Fine grained named entity recognition via seq2seq framework. IEEE Access, 8:53953– 53961. [↗](https://doi.org/10.1109/access.2020.2980431)

*Updated 1 November 2026*