CSKB-Population

CSKB-Population数据集是由香港科技大学计算机科学与工程系创建，旨在评估常识知识库的填充任务。该数据集通过整合四个流行的常识知识库（ConceptNet, ATOMIC, ATOMIC20 20, GLUCOSE）与大规模自动提取的事件图ASER，提供了一个高质量的人工标注评估集。数据集包含31700条高质量三元组，用于探测神经模型的常识推理能力。此数据集的应用领域主要集中在提升机器对未见声明的常识推理能力，解决现有评估方法的不准确性和规模限制问题。

The CSKB-Population dataset was developed by the Department of Computer Science and Engineering, The Hong Kong University of Science and Technology, with the goal of evaluating the common sense knowledge base population task. This dataset integrates four prominent common sense knowledge bases (ConceptNet, ATOMIC, ATOMIC 2020, GLUCOSE) with the large-scale automatically extracted event graph ASER, serving as a high-quality manually annotated evaluation benchmark. It contains 31,700 high-quality triples designed to probe the common sense reasoning capabilities of neural models. The primary application scenarios of this dataset focus on enhancing machines' common sense reasoning abilities toward unseen statements, and addressing the inaccuracy and scale limitation issues of existing evaluation methods.