GCite数据集

承担单位北京大学在项目执行过程中研究了解耦合论文引用关系预测模型，构建了大规模论文引用数据集，包含超过4.8K篇论文和25K条异质论文引用关系，课题组在该数据集基础上进行异质图表征和推理的相关研究，包括可解释解耦合推理、图谱表征等。数据集存储规模约11MB。Peking University, the implementing institution of the project, developed a decoupled paper citation relationship prediction model during the project implementation phase, and constructed a large-scale paper citation dataset. The dataset contains over 4.8 thousand papers and 25 thousand heterogeneous paper citation relations. Based on this dataset, the research team carried out relevant studies on heterogeneous graph representation and reasoning, including interpretable decoupled reasoning, graph representation, and other related research directions. The storage size of the dataset is approximately 11 MB.

搜集汇总

背景与挑战

背景概述

GCite数据集由北京大学构建，包含超过4.8K篇论文和25K条异质论文引用关系，存储规模约11MB。该数据集用于研究解耦合论文引用关系预测模型，支持异质图表征和推理等任务，如可解释解耦合推理和图谱表征。