Muradif

Muradif (مُرادِف, "synonym") is a synonym-based benchmark that directly assesses embedding quality with no additional fine-tuning. Each row is a triplet with a context: a model should embed context with anchor_word closer to context with syn_word (a true synonym) than to context with irrelevant_word (an unrelated word). There are 38,554 triplets in this benchmark.

This dataset was introduced at the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026). For more information visit our website: https://acr.ps/neoarabert.

Example row:

{
  "context": "يؤدي المسلمون صلاة الجمعة في ال_ جماعة",
  "anchor_word": "مَسْجِدٌ",
  "syn_word": "جَامِعٌ",
  "irrelevant_word": "اِنْجَابَ"
}

Citation

If you use this dataset, please cite:

@inproceedings{abou-chakra-etal-2026-neoarabert,
  title = "{NeoAraBERT}: A Modern Foundation Model for Arabic Embeddings with Diacritics-Aware Tokenization and POS-Targeted Masking",
  author = "Abou Chakra, Chadi and
            Hamoud, Hadi and
            Rakan Al Mraikhat, Osama and
            Abu Obaida, Qusai and
            Ballout, Mohamad and
            Zaraket, Fadi A.",
  booktitle = "Findings of the Association for Computational Linguistics: ACL 2026",
  address = "San Diego, California, United States",
  year = "2026",
  note = "Accepted paper",
  url = "https://acr.ps/neoarabert",
  abstract = {We present NeoAraBERT, a state-of-the-art open-source Arabic text-embedding model built on the NeoBERT architecture. We pre-train NeoAraBERT on diverse open-source and internal datasets covering modern standard, classical, and dialectal Arabic. We guided our design choices with Arabic tailored ablation studies including text normalization, light stemming, and diacritics-aware tokenization handling. We also performed more general POS-aware token masking and learning-rate scheduling ablation studies. We benchmarked NeoAraBERT against five top-performing Arabic models on 23 tasks, including a novel synonym-based task, ``Muradif'', that directly assesses embedding quality with no additional fine-tuning. NeoAraBERT variants (MSA, dialectal, and mixed) rank first in 18 tasks, second in two, third in two, and fourth in one task. They show strong performance on classical and modern standard Arabic, substantial margins of improvement ($>$7\%) in two tasks, and a $+$2.75\% improvement on average across all tasks. Our code and links to checkpoints for our model variants are available on our website: \url{https://acr.ps/neoarabert}}
}

License

This dataset is licensed under the CC BY-SA 4.0 license. The text of the license can be found here.