skill-eval-cranfield

name	skill-eval-cranfield
description	基于 Cranfield 信息检索范式的 skill 推荐系统评估方法论。涵盖分层测试查询设计(L1-L5)、自动 Qrels 生成、量化指标(Recall/MRR/NDCG/Spurious)、基线对比、按难度分层报告。适用于需要量化评估推荐/匹配系统改进效果的场景。
version	1.0.0
author	Emma (小喵)
license	MIT
triggers	["eval","评估","recall","mrr","ndcg","baseline","基线","ground truth","qrels","相关性判断","测试框架","cranfield","评估指标","量化测试","对照测试"]
metadata	{"hermes":{"tags":["evaluation","testing","information-retrieval","cranfield-paradigm","recommendation-system","benchmark"],"category":"testing","skill_type":"methodology","design_pattern":"pipeline","related_skills":["skill-creator","systematic-debugging"]}}
category	testing

skill-eval-cranfield

基于 Cranfield 信息检索范式的 skill 推荐系统评估方法论。涵盖分层测试查询设计(L1-L5)、自动 Qrels 生成、量化指标(Recall/MRR/NDCG/Spurious)、基线对比、按难度分层报告。适用于需要量化评估推荐/匹配系统改进效果的场景。