Historical Arabic Handwritten Text Images with Transcriptions

A collection of historical Arabic text spanning different geographies across centuries. Experts have transcribed forty historical pages, each five from a distinct book, providing textual ground truth for each image. The dataset was created by sherif1313 and last updated on February 14, 2026.

Use Cases

• Train OCR models for historical Arabic handwriting based on the provided image-text pairs.
• Test the performance of deep learning OCR models on diverse historical Arabic scripts.
• Analyze the evolution of Arabic handwriting styles across different centuries and geographies.

Strengths

• Contains forty meticulously transcribed historical pages.
• Includes five pages from each of eight distinct historical books, suggesting source diversity.

Limitations

• Row count and dataset size are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

huggingface

Collection Method

Expert transcription of historical pages from distinct books.

Time Range

Spans different centuries (specific range unknown).

Freshness

Last updated 2026-02-14 20:11:17; freshness should be verified.

Geography

Spans different geographies (specific locations unknown).