{
 "id": "MohammadJRanjbar/PersianPunc",
 "owner": "MohammadJRanjbar",
 "name": "PersianPunc",
 "downloads": 9,
 "likes": 5,
 "last_modified": "2026-04-12T09:30:41.000Z",
 "stated_licence": "cc-by-4.0",
 "stated_from": "tag",
 "licence_name": null,
 "licence_link": null,
 "file_licence": null,
 "licence_files": [],
 "sources": [],
 "gated": "auto",
 "tasks": [
  "token-classification",
  "text2text-generation"
 ],
 "citation": "@inproceedings{kalahroodi-etal-2026-persianpunc,\n    title = \"{P}ersian{P}unc: A Large-Scale Dataset and {BERT}-Based Approach for {P}ersian Punctuation Restoration\",\n    author = \"Kalahroodi, Mohammad Javad Ranjbar  and\n      Faili, Heshaam  and\n      Shakery, Azadeh\",\n    editor = \"Merchant, Rayyan  and\n      Megerdoomian, Karine\",\n    booktitle = \"The Proceedings of the First Workshop on {NLP} and {LLM}s for the {I}ranian Language Family\",\n    month = mar,\n    year = \"2026\",\n    address = \"Rabat, Morocco\",\n    publisher = \"Association for Computational Linguistics\",\n    url = \"https://aclanthology.org/2026.silkroadnlp-1.11/\",\n    doi = \"10.18653/v1/2026.silkroadnlp-1.11\",\n    pages = \"105--113\",\n    ISBN = \"979-8-89176-371-5\",\n    abstract = \"Punctuation restoration is essential for improving the readability and downstream utility of automatic speech recognition (ASR) outputs, yet remains underexplored for Persian despite its importance. We introduce PersianPunc, a large-scale, high-quality dataset of 17 million samples for Persian punctuation restoration, constructed through systematic aggregation and filtering of existing textual resources. We formulate punctuation restoration as a token-level sequence labeling task and fine-tune ParsBERT to achieve strong performance. Through comparative evaluation, we demonstrate that while large language models can perform punctuation restoration, they suffer from critical limitations: over-correction tendencies that introduce undesir",
 "signal": "permissive",
 "reason": "The declared licence (cc-by-4.0) is classified as permissive.",
 "revision": "c7dfd680233c5b2da55d38cc3acc2d7a49206b32",
 "stated_evidence": "https://huggingface.co/datasets/MohammadJRanjbar/PersianPunc/blob/c7dfd680233c5b2da55d38cc3acc2d7a49206b32/README.md",
 "file_evidence": "https://huggingface.co/datasets/MohammadJRanjbar/PersianPunc/tree/c7dfd680233c5b2da55d38cc3acc2d7a49206b32",
 "card_nc": false
}