Schema A
细颗粒度 RAG 片段
{"id":"ch01-kp-01","chapter":1,"section_type":"knowledge_point","text":"Four Controversies\nTrade war, immigration, Brexit and exchange-rate disputes set the analytical agenda."}适合按章节、知识点、案例、习题、答案、教材段落建立向量索引,做章节级检索与问答。
Dataset Shelf
这里集中放置课程主语料、细颗粒度切片、章节讲义和 chat fine-tuning 样本。适合做向量检索、知识助手或题库微调。
包含 25 章概览、统计信息与下载入口的总清单
用于前端检索和轻量级语义浏览的章节索引
按知识点、案例、习题、答案与教材段落切分的 JSONL
基于题库问答构造的 chat fine-tuning 样本
Schema A
{"id":"ch01-kp-01","chapter":1,"section_type":"knowledge_point","text":"Four Controversies\nTrade war, immigration, Brexit and exchange-rate disputes set the analytical agenda."}适合按章节、知识点、案例、习题、答案、教材段落建立向量索引,做章节级检索与问答。
Schema B
{"messages":[{"role":"system","content":"你是国际经济学课程助教..."},{"role":"user","content":"Chapter 8: Analysis of a Tariff..."},{"role":"assistant","content":"正确答案是 C..."}]}适合直接进入 chat fine-tuning 工作流,训练课程助教对题库问题做结构化回答。