ibm-research/REAL-MM-RAG_TechSlides_BEIR

这是一个与BEIR兼容的 REAL-MM-RAG_TechSlides 数据集版本，专为评估检索设置而重新格式化。原数据集是QA风格的，每个条目都是一个与文档图像相关联的查询。该数据集包含62个关于商业和IT自动化的技术演示文稿，共1963页，由文本、视觉和表格组成，用于评估从幻灯片演示文稿中检索IT自动化和商业洞察能力。

This is the BEIR-compatible version of the REAL-MM-RAG_TechSlides dataset, reformatted for evaluation in retrieval settings. The original dataset is QA-style, with each entry being a query tied to a document image. The dataset contains 62 technical presentations on business and IT automation, totaling 1,963 pages, composed of text, visuals, and tables, designed to evaluate the retrieval of IT automation and business insights from slide decks.