dwb2023/gdelt-rag-evaluation-datasets

GDELT RAG评估数据集是一个包含5种不同检索策略在GDELT RAG系统上测试的RAGAS评估输入数据集的集合。每种策略都在同一个由12个问题组成的黄金测试集上进行评估，从而直接比较检索性能。数据集包含约1400多条评估记录，涵盖了基准、朴素、BM25、集成和Cohere重排等检索策略。该数据集旨在用于基准测试RAG检索策略、比较不同的检索方法、分析检索质量以及训练检索模型。

The GDELT RAG Evaluation Datasets is a collection of RAGAS evaluation input datasets from 5 different retrieval strategies tested on the GDELT RAG system. Each strategy is evaluated on the same golden testset of 12 questions, providing a direct comparison of retrieval performance. The dataset contains approximately 1,400 evaluation records across baseline, naive, BM25, ensemble, and Cohere rerank retrieval strategies. It is intended for benchmarking RAG retrieval strategies, comparing different retrieval approaches, analyzing retrieval quality, and training retrieval models.