知识卡片
摘要相关性与完整性分母不同
相关性关注摘要中有多少内容值得保留,完整性关注源文重要信息有多少被涵盖。需要先定义句子、事实项及重要性,不能混用不同单位直接算精确率和召回率;摘要长度预算也影响可达完整性。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 外部工具与智能体
原书出处 · 1 处
- · 查看原书依据
Relevance Also called precision, this is calculated as the percentage of sentences in the summary that are deemed relevant and thus merit inclusion in the summary. Completeness Also called recall, this is calculated as the percentage of relevant items in the source document that are included in the summary. Repetitiveness The summary should not be repetitive, even if there is repetition in the source document.
参考来源
- 位置:《Designing Large Language Model Applications》第10章《外部工具与智能体》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Designing Large Language Model Applications/_epub-src/OEBPS/ch10.html)
- 结论依据:摘录中关于摘要相关性与完整性分母不同的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:Relevance Also called precision, this is calculated as the percentage of sentences in the summary that are deemed relevant and thus merit inclusion in the summary. Completeness Also called recall, this is calculated as the percentage of relevant items in the source document that are included in the summary. Repetitiveness The summary should not be repetitive, even if there is repetition in the source document.