D-Nikud: Enhancing Hebrew Diacritization with LSTM and Pretrained Models

Adi Rosenthal, Nadav Shaked
1/30/2024
cs.CL

No official code found. Best implementation: NadavShaked/D_Nikud (3 stars, last commit 2025-12-07).

Abstract

D-Nikud, a novel approach to Hebrew diacritization that integrates the strengths of LSTM networks and BERT-based (transformer) pre-trained model. Inspired by the methodologies employed in Nakdimon, we integrate it with the TavBERT pre-trained model, our system incorporates advanced architectural choices and diverse training data. Our experiments showcase state-of-the-art results on several benchmark datasets, with a particular emphasis on modern texts and more specified diacritization like gender.

View on arXivView PDF

Code Implementations(1)

NadavShaked/D_NikudCommunity100%
32PythonJul 19, 202310 months ago

Similar Papers

MenakBERT -- Hebrew Diacriticizer

84% match2024

Dicta-LM 3.0: Advancing The Frontier of Hebrew Sovereign LLMs

81% match2026

NeoDictaBERT: Pushing the Frontier of BERT models for Hebrew

81% match2025

Adapting LLMs to Hebrew: Unveiling DictaLM 2.0 with Enhanced Vocabulary and Instruction Capabilities

80% match2024

HalleluBERT: Let every token that has meaning bear its weight

80% match2025

Are LLMs Good Text Diacritizers? An Arabic and Yorùbá Case Study

77% match2025

Cite this paper

@article{rosenthal2024dnikud,
  title  = {D-Nikud: Enhancing Hebrew Diacritization with LSTM and Pretrained Models},
  author = {Adi Rosenthal and Nadav Shaked},
  year   = {2024},
  eprint = {2402.00075},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2402.00075v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2402.00075.svg)](https://codeofpaper.com/paper/2402.00075)

Discussion