RiceChem
收藏资源简介:
RiceChem数据集是由莱斯大学开发的一个专门用于自动长答案评分(ALAG)研究的数据集。该数据集包含1264个来自大学化学课程的学生长答案响应,每个响应都针对27个评分标准进行评分,总计形成8392个数据点。RiceChem的平均字数为120,远高于其他ASAG数据集,这使得它非常适合用于探索ALAG的复杂性。数据集的创建过程涉及多个教学助理对学生响应进行评分,使用TRUE或FALSE标签来标记每个评分标准是否被正确回答。RiceChem数据集的应用领域主要集中在教育领域,旨在通过自动化技术提高长答案评分的准确性和效率,从而为教育评估提供更可靠的工具。
The RiceChem dataset was developed by Rice University for research on Automated Long Answer Grading (ALAG). This dataset contains 1,264 long student responses from college chemistry courses, each of which is scored against 27 grading criteria, resulting in a total of 8,392 data points. The RiceChem dataset has an average word count of 120 per response, which is significantly higher than that of other ASAG datasets, making it highly suitable for exploring the complexity of ALAG. The creation of the RiceChem dataset involved multiple teaching assistants scoring student responses, with TRUE or FALSE labels used to indicate whether each grading criterion was correctly addressed. The primary application domain of the RiceChem dataset is education, where it aims to enhance the accuracy and efficiency of long answer grading via automated technologies, thus providing more reliable tools for educational assessment.

- 1Automated Long Answer Grading with RiceChem Dataset莱斯大学 · 2024年



