whybe-choi/kovidore-office-v1.0-beir

KoViDoRe - Office v1.0

This dataset is part of the KoViDoRe Benchmark for Korean visual document retrieval evaluation. Specifically, it is focused on Office Documents such as reports and press releases for text-to-image retrieval tasks.

Dataset Summary

• Domain: Korean business/office documents (e.g., 보고서, 보도자료 등)
• Task: Visual document retrieval (text query → relevant document image)
• Format: BEIR-compatible (corpus, queries, qrels)

Statistics:

• Number of Documents (Pages): 1,990+
• Number of Queries: 222
• Number of Relevance Judgments (qrels): 244
• All text queries are in Korean.

Dataset Structure

• corpus
  • corpus-id: Unique page ID
  • image: Document page image (PIL Image)
  • title: Document title or file name
• queries
  • query-id: Unique query ID
  • query: Korean text question
• qrels
  • query-id
  • corpus-id
  • score: (1 = relevant)

Usage

This dataset is intended for benchmarking the performance of visual document retrieval systems on real-world Korean reports and press releases.

Note:
Due to licensing restrictions, original images are not redistributed with this dataset.
Please download the required image data directly from AI Hub (오피스 문서 생성 데이터) and place them as instructed in the KoVidore-benchmark GitHub repository.

Acknowledgments

• Based on data provided by AI Hub (오피스 문서 생성 데이터)

---