Research

Paper

AI LLM March 23, 2026

DATASHI: A Parallel English-Tashlhiyt Corpus for Orthography Normalization and Low-Resource Language Processing

Authors

Nasser-Eddine Monir, Zakaria Baou

Abstract

DATASHI is a new parallel English-Tashlhiyt corpus that fills a critical gap in computational resources for Amazigh languages. It contains 5,000 sentence pairs, including a 1,500-sentence subset with expert-standardized and non-standard user-generated versions, enabling systematic study of orthographic diversity and normalization. This dual design supports text-based NLP tasks - such as tokenization, translation, and normalization - and also serves as a foundation for read-speech data collection and multimodal alignment. Comprehensive evaluations with state-of-the-art Large Language Models (GPT-5, Claude-Sonnet-4.5, Gemini-2.5-Pro, Mistral, Qwen3-Max) show clear improvements from zero-shot to few-shot prompting, with Gemini-2.5-Pro achieving the lowest word and character-level error rates and exhibiting robust cross-lingual generalization. A fine-grained analysis of edit operations - deletions, substitutions, and insertions - across phonological classes (geminates, emphatics, uvulars, and pharyngeals) further highlights model-specific sensitivities to marked Tashlhiyt features and provides new diagnostic insights for low-resource Amazigh orthography normalization.

Metadata

arXiv ID: 2603.21571
Provider: ARXIV
Primary Category: cs.CL
Published: 2026-03-23
Fetched: 2026-03-24 06:02

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.21571v1</id>\n    <title>DATASHI: A Parallel English-Tashlhiyt Corpus for Orthography Normalization and Low-Resource Language Processing</title>\n    <updated>2026-03-23T04:46:59Z</updated>\n    <link href='https://arxiv.org/abs/2603.21571v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.21571v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>DATASHI is a new parallel English-Tashlhiyt corpus that fills a critical gap in computational resources for Amazigh languages. It contains 5,000 sentence pairs, including a 1,500-sentence subset with expert-standardized and non-standard user-generated versions, enabling systematic study of orthographic diversity and normalization. This dual design supports text-based NLP tasks - such as tokenization, translation, and normalization - and also serves as a foundation for read-speech data collection and multimodal alignment. Comprehensive evaluations with state-of-the-art Large Language Models (GPT-5, Claude-Sonnet-4.5, Gemini-2.5-Pro, Mistral, Qwen3-Max) show clear improvements from zero-shot to few-shot prompting, with Gemini-2.5-Pro achieving the lowest word and character-level error rates and exhibiting robust cross-lingual generalization. A fine-grained analysis of edit operations - deletions, substitutions, and insertions - across phonological classes (geminates, emphatics, uvulars, and pharyngeals) further highlights model-specific sensitivities to marked Tashlhiyt features and provides new diagnostic insights for low-resource Amazigh orthography normalization.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CL'/>\n    <published>2026-03-23T04:46:59Z</published>\n    <arxiv:comment>This paper has been accepted for presentation at LREC 2026</arxiv:comment>\n    <arxiv:primary_category term='cs.CL'/>\n    <author>\n      <name>Nasser-Eddine Monir</name>\n    </author>\n    <author>\n      <name>Zakaria Baou</name>\n    </author>\n  </entry>"
}