KOLongDoc: Korean Long Document Benchmark for Multimodal and RAG Evaluation

A benchmark dataset for evaluating AI models on Korean long and complex documents, created by Markr-AI. It contains 136 'Long Document Problems' and 64 'Super Long Document Problems', as described on the dataset page. The dataset was last updated on 2026-05-30.

Use Cases

• Benchmarking Korean LLM performance on long-document comprehension based on the described document problems.
• Evaluating Retrieval-Augmented Generation (RAG) systems for Korean administrative workflows based on the described complex document tasks.
• Training or fine-tuning multimodal models for Korean document understanding based on the dataset's described scope.
• Assessing model capabilities on super-long Korean texts based on the separate category of 64 problems.

Strengths

• Contains 136 specifically defined 'Long Document Problems' for evaluation.
• Includes a separate subset of 64 'Super Long Document Problems' for more challenging tasks.
• Addresses a noted gap in Korean-language long-document benchmarks as described.

Limitations

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

Provenance

Source

Markr-AI via Hugging Face.

Collection Method

Method of collection is not specified in the provided input.

Freshness

Last updated 2026-05-30 16:00:37; freshness should be verified.

Geography

South Korea (inferred from focus on Korean language and administrative tasks).