Melady/TSAIA

TSAIA是一个专门为时间序列分析和多项选择题任务设计的教学评估数据集，包含两个子集：analysis_questions和multiple_choice。analysis_questions子集包含904个样本，每个样本包含问题ID、问题类型、任务的自然语言描述、嵌入式时间序列数据、执行变量定义、参考答案或目标输出、任务上下文以及输出格式或变量命名的约束等字段。multiple_choice子集包含150个样本，字段信息与analysis_questions类似，但多了选项和正确答案。

TSAIA is an instructional assessment dataset specifically designed for time series analysis and multiple choice tasks, containing two subsets: analysis_questions and multiple_choice. The analysis_questions subset includes 904 samples, with fields such as question ID, question type, natural language description of the task, embedded time series data, definition of executor variables, reference answer or target output, context of the task, and constraints on output format or variable naming. The multiple_choice subset contains 150 samples, with similar fields to analysis_questions but also including options and correct answers.