---
title: "Credibility Assessment for Arabic News on the Gaza War: A"
description: "The paper addresses automatic credibility checking of Arabic news about the Gaza war. 21 references."
authors: ["Sanaa Abril", "Sihame Mouanid", "El habib Ben lahmar", "Omar Zahour"]
year: 2026
venue: "Proceedings of the 2nd International Workshop on Nakba Narratives as Language Resources @ LREC 2026"
source: "https://aclanthology.org/2026.nakbanlp-1.5/"
doi: "10.63317/2hvdmcb8xkoh"
licence: "CC BY-NC 4.0 (ELRA, LREC proceedings)"
licence_url: "https://creativecommons.org/licenses/by-nc/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/abril-2026-credibility-assessment-arabic-news-gaza/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/abril-2026-credibility-assessment-arabic-news-gaza/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/26/) / Credibility Assessment for Arabic News on the Gaza War: A Hybrid Neural-Symbolic Pipeline

# Credibility Assessment for Arabic News on the Gaza War: A Hybrid Neural-Symbolic Pipeline

**Authors:** Sanaa Abril, Sihame Mouanid, El habib Ben lahmar and Omar Zahour · **Year:** 2026 · **Venue:** [Proceedings of the 2nd International Workshop on Nakba Narratives as Language Resources @ LREC 2026](https://md.research.elamur.ai/free/en/venues/nakbanlp/) ([International Workshop on Nakba Narratives as Language Resources @ LREC 2026](https://md.research.elamur.ai/free/en/venues/nakbanlp/2026/)) · **Publisher:** [European Language Resources Association (ELRA)](https://md.research.elamur.ai/free/en/publishers/european-language-resources-association-elra/) · **Licence:** [CC BY-NC 4.0 (ELRA, LREC proceedings)](https://creativecommons.org/licenses/by-nc/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.nakbanlp-1.5/) · [PDF](https://aclanthology.org/2026.nakbanlp-1.5.pdf) · [DOI 10.63317/2hvdmcb8xkoh](https://doi.org/10.63317/2hvdmcb8xkoh) · [OpenAlex](https://openalex.org/W7166590746)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
The paper addresses automatic credibility checking of Arabic news about the Gaza war. In this domain, emotionally loaded framing and coordinated disinformation make it hard to build trustworthy archives of conflict records. The authors assembled a binary real/fake corpus of 16,708 instances from three sources: the AFND benchmark, Gaza-related posts on X labeled by source rules (bot or inauthentic accounts, fake-news hashtags), and articles from established outlets treated as real. No manual annotation was done. The corpus was filtered with conflict keywords and limited to Modern Standard Arabic. Their pipeline fuses three signals through a gated fusion step: frozen AraBERT sentence vectors, a CNN that captures local n-gram patterns, and an interpretable fuzzy score for sensational vocabulary. They then compare classical ML classifiers, deep models and LLM baselines. BiLSTM performed best, AraBERT came very close, and general-purpose LLMs fell well behind. The best model was released as a browser plugin that shows a label, a confidence score and highlighted rationale spans. The authors caution that source-based labeling may inflate the reported scores.

## Key points
- Data sources: AFND (original labels kept), keyword-collected X posts about the Gaza and Palestine–Israel conflict, and Al Jazeera Arabic and Al Arab articles as the real class. The corpus has 16,708 labeled instances, MSA only, with no dialects.
- AraBERT is kept frozen. Final-layer states are mean-pooled over non-padding tokens to give one 768-dimensional vector per text.
- The CNN branch uses kernel sizes {3, 4, 5} with 100 filters each, ReLU and global max pooling, producing a 300-dimensional vector. This is fused with the AraBERT vector and the fuzzy EWMF score.
- Results: BiLSTM reached Accuracy/F1 = 0.9867 and Recall = 0.9912, while AraBERT reached Accuracy/F1 = 0.9838. Classical ML baselines were strong but lower, and LLM baselines lagged substantially.
- The deployed plugin explains predictions with Integrated Gradients rationales. Spans are kept only if deleting them noticeably lowers confidence.

## Interesting ideas
- Instead of a hard yes/no lexicon flag, the Extravagant Word Membership Function turns counts of exaggeration words, intensifiers and clickbait patterns into a graded fuzzy degree, giving one transparent, tunable feature beside dense vectors.
- Rationales from Integrated Gradients are not trusted as-is; a span survives only if removing it measurably drops the model's confidence, a cheap faithfulness check on attributions.
- The authors themselves note that labels assigned by source (outlet vs. bot account) may let the classifier learn source or topic style rather than the target property, which makes very high scores suspect.
- The fusion design keeps a frozen pretrained encoder and combines it with a small local-pattern extractor and a hand-designed interpretable feature through a learned gate, so each signal remains inspectable.

## For Mova

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

- **I1** (frames): A similar graded hyperbole or intensity score from lexicon hits could feed the irony and implicature module as an explainable cue rather than a binary trigger.
- **I2** (learn): The same deletion test could validate which tokens or edges a learned parser or repair rule really depends on, filtering out spurious explanations.
- **I3** (learn): When silver data is labeled by provenance, add negative controls such as source-balanced splits and near-duplicate removal so that evaluation detects shortcut learning.
- **I4** (tree): The gparse transformer parser could gate frozen encoder features together with explicit symbolic cues (morphology, lexicon flags), keeping their contribution measurable per decision.

*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 Hassan II Casablanca](https://md.research.elamur.ai/free/en/universities/university-of-hassan-ii-casablanca/) — University, Casablanca, Morocco · 2 works

📚 Shelves (5): [🖥️ 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 · [🗄️ Information Systems](https://md.research.elamur.ai/free/en/shelves/computer-science/information-systems/) · 1,437 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: Sanaa Abril, Sihame Mouanid, El habib Ben lahmar and Omar Zahour. 2026. Credibility Assessment for Arabic News on the Gaza War: A Hybrid Neural-Symbolic Pipeline. Proceedings of the 2nd International Workshop on Nakba Narratives as Language Resources @ LREC 2026.

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

```bibtex
@inproceedings{abril2026credibility,
  title = {{Credibility Assessment for Arabic News on the Gaza War: A Hybrid Neural-Symbolic Pipeline}},
  author = {Sanaa Abril, Sihame Mouanid, El habib Ben lahmar and Omar Zahour},
  booktitle = {Proceedings of the 2nd International Workshop on Nakba Narratives as Language Resources @ LREC 2026},
  year = {2026},
  doi = {10.63317/2hvdmcb8xkoh},
  url = {https://doi.org/10.63317/2hvdmcb8xkoh}
}
```

- [← previous Credal Concept Bottleneck Models for Epistemic–Aleatoric Uncertainty Decomposition](https://md.research.elamur.ai/free/en/papers/mukherjee-2026-credal-concept-bottleneck-models-epistemic/)
- [next → CreditDecoding: Accelerating Parallel Decoding in Diffusion Large Language Models with Trace Credit](https://md.research.elamur.ai/free/en/papers/wang-2026-creditdecoding-accelerating-parallel-decoding-diffusion/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/26/)

## References

21 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Maha Al-yahya, Hend Al-Khalifa, Heyam Al-Baity, Duaa AlSaeed, and A. Essam. 2021. Arabic fake news detection: Comparative study of neural networks and transformer-based approaches. [↗](https://doi.org/10.1155/2021/5516945)
2. Ahmad Alsharairi, Hana Abdul-Rahman Al-Souob, Mohammad Farouq AlQadi, and Sora Mohammad Shatnawi. 2024. Social media communication and framing of the gaza conflict: Impact on public opinion. ResearchGate. [↗](https://doi.org/10.36923/jicc.v25i3.1153)
3. Despoina Antonakaki and Sotiris Ioannidis. 2025. Cross-platform digital discourse analysis of the israel-hamas conflict: Sentiment, topics, and event dynamics. [↗](https://doi.org/10.48550/arxiv.2601.02367)
4. Widad Awane, El Habib Ben Lahmar, and Ayoub El Falaki. 2021. Hate speech in the arab electronic press and social networks. Revue d’Intelligence Artificielle, 35(6):457. [↗](https://doi.org/10.18280/ria.350603)
5. Dipto Barman, Ziyi Guo, and Owen Conlan. 2024. The dark side of language models: Exploring the potential of llms in multimedia disinformation generation and dissemination. Machine Learning with Applications, 16:100545. [↗](https://doi.org/10.1016/j.mlwa.2024.100545)
6. Mohammed Ennaouri and Ahmed Zellou. 2025. Enhancing fake review detection using linguistic exaggeration, bert embeddings, and fuzzy logic. IEEE Access, 13:135957–135968. [↗](https://doi.org/10.1109/access.2025.3594629)
7. Noureddine Ettaik and El Habib Ben Lahmar. 2021. Hyperparameter optimisation in nlp architectures. In Proceedings of the 2nd International Conference on Big Data, Modelling and Machine Learning, pages 55–62, Lisbon, Portugal. SCITEPRESS - Science and Technology Publications. [find](https://scholar.google.com/scholar?q=Noureddine+Ettaik+and+El+Habib+Ben+Lahmar%2E+2021%2E+Hyperparameter+optimisation+in+nlp+architectures%2E+In+Proceedings+of+the+2nd+International+Conference+on+Big+Data%2C+Modelling+and+Machine+Learning%2C+pages)
8. Khaled M. Fouad, Sahar F. Sabbeh, and Walaa Medhat. 2022. Arabic fake news detection using deep learning. Computers, Materials & Continua, 71(2):3647–3665. [↗](https://doi.org/10.32604/cmc.2022.021449)
9. Reto Gubelmann and Ghassen Karray. 2025. Assessing reliability and political bias in LLMs’ judgements of formal and material inferences with partisan conclusions. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 30005–30031, Vienna, Austria. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Reto+Gubelmann+and+Ghassen+Karray%2E+2025%2E+Assessing+reliability+and+political+bias+in+LLMs%E2%80%99+judgements+of+formal+and+material+inferences+with+partisan+conclusions%2E+In+Proceedings+of+the+63rd+Annual+Mee)
10. Suhaib Kh Hamed, Mohd Juzaiddin Ab Aziz, and Mohd Ridzwan Yaakub. 2023. A review of fake news detection approaches: A critical analysis of relevant studies and highlighting key challenges associated with the dataset, feature representation, and data fusion. Heliyon, 9(10):e20382. [↗](https://doi.org/10.1016/j.heliyon.2023.e20382)
11. Institute for Strategic Dialogue. 2025. Conflict Amplified: Disinformation and Hate in the Israel-Hamas War. Institute for Strategic Dialogue. [find](https://scholar.google.com/scholar?q=Institute+for+Strategic+Dialogue%2E+2025%2E+Conflict+Amplified%3A+Disinformation+and+Hate+in+the+Israel%2DHamas+War%2E+Institute+for+Strategic+Dialogue%2E)
12. Jumana Jouhar, Anju Pratap, Neharin Tijo, and Meenakshi Mony. 2024. Fake news detection using python and machine learning. Procedia Computer Science, 233:763–771. 5th International Conference on Innovative Data Communication Technologies and Application (ICIDCA 2024). [find](https://scholar.google.com/scholar?q=Jumana+Jouhar%2C+Anju+Pratap%2C+Neharin+Tijo%2C+and+Meenakshi+Mony%2E+2024%2E+Fake+news+detection+using+python+and+machine+learning%2E+Procedia+Computer+Science%2C+233%3A763%E2%80%93771%2E+5th+International+Conference+on+Innov)
13. Ashwaq Khalil, Moath Jarrah, Monther Aldwairi, and Manar Jaradat. 2022. Afnd: Arabic fake news dataset for the detection and classification of articles credibility. Data in Brief, 42:108141. [↗](https://doi.org/10.60692/7jqa1-sqb28)
14. Baqer M. Merzah, Jafar Razmara, and Zolfaghar Salmanian. 2025. Hybrid deep learning models for fake news detection: Case study on arabic and english languages. Frontiers in Big Data. [↗](https://doi.org/10.3389/fdata.2025.1683786)
15. Mahmud Mirza Saad, Shahzad Muhammad Khuram, Ali Syed Imran, Jabeen Farzana, and Moetesum Momina. 2025. Fond: Fuzzy-based optimized fake-news detection using big-bird and longformer. SSRN Electronic Journal. [↗](https://doi.org/10.2139/ssrn.5349353)
16. Ashraf Bany Mohammed, Manaf Al-Okaily, Dhia Qasim, Shafique Ur Rehman, and Latifa Abdalla. 2025. Digital activism and public opinion: understanding the role of social media during the gaza conflict. Journal of Islamic Marketing, 16(11):3366–3393. [↗](https://doi.org/10.1108/jima-03-2024-0101)
17. Eleftheria Papageorgiou, Christos Chronis, Iraklis Varlamis, and Yassine Himeur. 2024. A survey on the use of large language models (llms) in fake news. Future Internet, 16(8):298. [↗](https://doi.org/10.3390/fi16080298)
18. Shaina Raza, Drai Paulen-Patterson, and Chen Ding. 2025. Fake news detection: comparative evaluation of bert-like models and large language models with generative ai-annotated data. [↗](https://doi.org/10.1007/s10115-024-02321-1)
19. H. Sallami, E. Aïmeur, et al. 2025. Aletheia: Detect, discuss, and stay informed on fake news. In Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence (IJCAI-25), Demonstrations Track. [find](https://scholar.google.com/scholar?q=H%2E+Sallami%2C+E%2E+A%C3%AFmeur%2C+et+al%2E+2025%2E+Aletheia%3A+Detect%2C+discuss%2C+and+stay+informed+on+fake+news%2E+In+Proceedings+of+the+Thirty%2DFourth+International+Joint+Conference+on+Artificial+Intelligence+%28IJCAI%2D25%29%2C)
20. Hussain Mohammed Turki, Essam Al Daoud, Ghassan Samara, Raed Alazaidah, Mais Haj Qasem, Mohammad Aljaidi, Suhaila Abuowaida, and Nawaf Alshdaifat. 2025. Arabic fake news detection using hybrid contextual features. International Journal of Electrical and Computer Engineering (IJECE), 15(1):836–845. [↗](https://doi.org/10.11591/ijece.v15i1.pp836-845)
21. Mohammed Abbas Yousef, Abeer ElKorany, and Hanaa Bayomi. 2024. Fake-news detection: a survey of evaluation arabic datasets. Social Network Analysis and Mining, 14:225. [↗](https://doi.org/10.1007/s13278-024-01370-2)

*Updated 1 November 2026*