TACM12K

该数据集是从ACM异构图数据集中扩展的一个关系表数据集，包含四个表格：论文、作者、引用和著作。这些表格具备如年份、标题、摘要和作者隶属关系等特征。此外，该数据集还包括四个表格的丰富特征以及对部分缺失论文特征的补充。在规模上，每个类别有20个标注样本，500个验证样本以及1000个测试样本。当前的任务是预测论文的会议归属。

This dataset is a relational table dataset expanded from the ACM heterogeneous graph dataset, which includes four tables: Papers, Authors, Citations, and Authorship. These tables possess attributes such as year, title, abstract, and author affiliation. Additionally, the dataset provides rich features for all four tables, along with imputation for some missing paper attributes. In terms of dataset scale, each category contains 20 labeled samples, 500 validation samples, and 1000 test samples. The current task is to predict the conference affiliation of papers.