AI4SoL/KC_Tagging

这个数据集包含了来自多个来源（ASDiv，GSM8K，MATH）的数学单词问题，并带有知识组件的注释，统一到一个一致的结构中。数据集目前仍在建设中，GSM8K和MATH部分的注释数据尚未完成，未来的版本将会提高完整性。数据集划分为四个部分：ASDiv数据集（2002个样本，已完全注释）、GSM8K数据集的两个批次（共12535个样本，其中一个批次正在注释中）和MATH数据集（3675个样本，正在注释中）。

This dataset contains mathematical word problems annotated with knowledge components from multiple sources (ASDiv, GSM8K, MATH), unified into a consistent structure. The dataset is still under construction with the annotation data for the GSM8K and MATH splits not yet complete. Completeness will improve in subsequent versions. The dataset is split into four parts: the ASDiv dataset (2,002 samples fully annotated), two batches of the GSM8K dataset (12,535 samples with one batch ongoing annotation), and the MATH dataset (3,675 samples ongoing annotation).