# PersianPunc: A Large-Scale Dataset and BERT-Based Approach for Persian Punctuation Restoration

arXiv: https://arxiv.org/abs/2603.05314 · HTML: https://mlsanity.com/paper/2603.05314/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 23 references that are in the map.
- First posted on arXiv 5 Mar 2026.

## Builds on (most cited first, 7 of 7)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [ParsBERT: Transformer-based Model for Persian Language Understanding](https://mlsanity.com/paper/2005.12515/) (2020)
- [Punctuation Prediction Model for Conversational Speech](https://mlsanity.com/paper/1807.00543/) (2018)
- [Automatic punctuation restoration with BERT models](https://mlsanity.com/paper/2101.07343/) (2021)
- [ViraPart: A Text Refinement Framework for Automatic Speech Recognition and Natural Language Processing Tasks in Persian](https://mlsanity.com/paper/2110.09086/) (2021)

## Built on by (most cited first, 1 of 1)

- [The Annotation Bottleneck in Persian Text NLP: Persian as an Annotation-Scarce Language](https://mlsanity.com/paper/2608.24698/) (2026)

## Abstract (arXiv)

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 undesired edits beyond punctuation insertion (particularly problematic for speech-to-text pipelines) and substantially higher computational requirements. Our lightweight BERT-based approach achieves a macro-averaged F1 score of 91.33% on our test set while maintaining efficiency suitable for real-time applications. We make our dataset (https://huggingface.co/datasets/MohammadJRanjbar/persian-punctuation-restoration) and model (https://huggingface.co/MohammadJRanjbar/parsbert-persian-punctuation) publicly available to facilitate future research in Persian NLP and provide a scalable framework applicable to other morphologically rich, low-resource languages.