ConcurrentQA Benchmark

该数据集名为ConcurrentQA，是首个要求在多个数据分布上同时检索的文本问答基准，它既反映了公共数据也包含了私有数据。该数据集允许在PAIR框架下评估多跳开放域问答（ODQA）系统，同时解决了隐私与性能之间的权衡问题。其所涉及的任务为问答（QA）。

This dataset, named ConcurrentQA, is the first text-based question answering benchmark that requires simultaneous retrieval across multiple data distributions, and incorporates both public and private data. It enables the evaluation of multi-hop open-domain question answering (ODQA) systems under the PAIR framework, while addressing the trade-off between privacy and performance. The task involved in this dataset is question answering (QA).