三个互补数据集: 专业目录(分类)、专业 grounding(深度知识)、院校索引(层次/批次)。

Three complementary datasets: the major catalogue (taxonomy), the major grounding (deep knowledge), and the school index (tier/batch).

三个数据集 / The three datasets

中文。

文件内容规模
major_catalogue.json完整本科专业目录,按学科门类组织(by_menlei:历史学/经济学/法学…→专业列表)全目录
major_grounding.json每个专业的深度知识:facts(核心课程·典型去向·真实岗位,公开报告级)+ lived(日常·真实的坑·常见转向·圈内评价,经验判断),事实/观点分别标注856 专业(859 张卡)
school_index_sample.json院校索引:每校 {985, 211, 双一流, 录取批次}2347 校

关键边界:院校索引只有层次与批次,没有录取分/位次——那属于 [data-admissions]

English.

filecontentsize
major_catalogue.jsonthe full 本科专业目录, organized by 学科门类 (by_menlei: 历史学/经济学/法学… → major lists)full catalogue
major_grounding.jsonper-major deep knowledge: facts (core courses · typical paths · real jobs, report-level) + lived (daily reality · real pitfalls · common pivots · in-field reputation, experiential), fact/opinion labeled859 majors
school_index_sample.jsonthe school index: {985, 211, 双一流, 录取批次} per school2347 schools

Key boundary: the school index carries only tier & batchnot admission scores/rank, which live in data-admissions (from 雨来).

怎么被用 / How it’s used

中文。 grounding.pygrounding_block(label)major_grounding 把专业落到真实课程/岗位/转向(对抗”泛泛冷门专业”的空洞);院校 985/211/双一流 标志用于分层与对比。机制见 grounding-layer

English. grounding_block(label) in grounding.py uses major_grounding to ground a major in real courses/jobs/pivots (countering generic “obscure major” emptiness); the 985/211/双一流 flags feed tiering and contrast. Mechanism: grounding-layer.

data-overview · data-employment · data-admissions · grounding-layer