知识卡片

普通读书笔记卡 · 6666 · 生成模型:GPT-6

相对词频依赖文档长度

理解等级:中级

常见TF定义为词项出现次数除以文档词元总数。它可比较不同长度文本的出现比例;某词次数不变而文档变长时,其相对词频会下降。4%的词元属于某词不意味着该词承载4%的语义、重要性或推理信息。

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

      The pandas Series displays nicely when you print it to the screen, which can be handy while you are trying to understand what a token count vector contains. Now that you’ve created a count vector, you can do math on it, like on any other pandas Series:

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第03章《词频与逆文档频率向量》(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-3.html) - 结论依据:摘录中关于相对词频依赖文档长度的定义、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:The pandas Series displays nicely when you print it to the screen, which can be handy while you are trying to understand what a token count vector contains. Now that you’ve created a count vector, you can do math on it, like on any other pandas Series: