RUKOPYS: Ukrainian Handwritten Text Recognition Dataset Spanning a Century

RUKOPYS is the first large-scale open dataset for Ukrainian handwritten text recognition. It spans over a century of Ukrainian handwriting, from 1920s archival documents to present-day school homework, and is designed for end-to-end document understanding tasks. The dataset was created by the UkrainianCatholicUniversity and was last updated on 2026-04-16.

Use Cases

• Train handwritten text recognition models based on the described collection of Ukrainian handwriting samples.
• Develop models for document region detection based on the dataset's design for end-to-end document understanding.
• Conduct historical linguistic analysis based on handwriting samples spanning from the 1920s to the present.
• Build educational tools for Ukrainian language learning based on the inclusion of present-day school homework samples.

Strengths

• First large-scale open dataset for Ukrainian handwritten text recognition.
• Spans over a century of handwriting, providing temporal diversity.
• Designed for multiple document understanding tasks: region detection, type classification, and text transcription.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.
• Data may reflect temporal and source bias inherent to the collected documents.

Provenance

Source

UkrainianCatholicUniversity

Collection Method

Likely collected from archival documents and contemporary sources such as school homework.

Time Range

1920s to present day

Freshness

Last updated 2026-04-16 07:31:42; freshness should be verified.

Geography

Ukraine (implied by the focus on the Ukrainian language)