← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6764.a · 生成模型:GPT-6

倒排索引加速候选定位非全程常数

理解等级:中级

倒排表从词项定位出现它的文档或位置,避免每次扫描全部文本。字典查找可以很快,但读取长posting、合并候选和排序仍有成本;不能把整个查询一概写成O(1),短语与词形还需对应索引设计。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Full-text searches started to appear very quickly, due to people’s need to find information related to keywords. Indexing, and especially inverted indexing, was what helped this search to be fast and efficient. Inverted indexes work similarly to the way you would find a topic in a textbook—by looking at the index at the end of the book and finding the page numbers where the topic

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第10章《真实世界中的大语言模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Natural Language Processing in Action, Second Edition/_epub-src/OEBPS/Text/chapter-10.html) - 结论依据:摘录中关于倒排索引加速候选定位非全程常数的定义、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:Full-text searches started to appear very quickly, due to people’s need to find information related to keywords. Indexing, and especially inverted indexing, was what helped this search to be fast and efficient. Inverted indexes work similarly to the way you would find a topic in a textbook—by looking at the index at the end of the book and finding the page numbers where the topic