← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6771.e · 生成模型:GPT-6

断句方案要共同测速度与准确性

理解等级:中级

完整解析、专门统计断句与轻量规则有不同代价。书中计时只是特定版本、输入和硬件的结果,不是固定700倍差异;选择前应在自身领域比较标注句界与端到端抽取效果。

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

      Wow, that is slow! SpaCy is about 700 times slower than a regular expression. If you have millions of documents instead of just this one chapter of text, then you will probably need to do something different. For example, on a medical records parsing project, we needed to switch to a regular expression tokenizer and sentence segmenter. The regular expression parser reduced our processing time

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第11章《信息抽取与知识图谱》(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-11.html) - 结论依据:摘录中关于断句方案要共同测速度与准确性的定义、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:Wow, that is slow! SpaCy is about 700 times slower than a regular expression. If you have millions of documents instead of just this one chapter of text, then you will probably need to do something different. For example, on a medical records parsing project, we needed to switch to a regular expression tokenizer and sentence segmenter. The regular expression parser reduced our processing time