知识卡片

结构图卡 · 6782 · 生成模型:GPT-6

知识抽取由多个可错环节协作

理解等级:中级

结构图:

classDiagram
 class 文本来源 {版本及原文位置}
 class 分句与语言注释 {边界词性及依存}
 class 实体识别 {提及跨度与类型}
 class 共指与链接 {提及链与稳定实体ID}
 class 关系抽取 {角色关系及限定条件}
 class 知识库 {主张来源与冲突}
 文本来源 --> 分句与语言注释
 分句与语言注释 --> 实体识别
 实体识别 --> 共指与链接
 共指与链接 --> 关系抽取
 分句与语言注释 --> 关系抽取
 关系抽取 --> 知识库 : 核验后写入

各步骤都需保存到原文的映射,后一步成功不自动抵消前一步误标。

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

      That’s just a simple use case. You can construct a graph from scratch using your own commonsense knowledge or the domain knowledge you want your AI to know about. But if you can extract knowledge from text, you can build much larger knowledge graphs much more quickly. Plus, you will need this algorithm to double-check any text generated by your language models. Knowledge extraction requires

    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:摘录中关于知识抽取由多个可错环节协作的定义、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:That’s just a simple use case. You can construct a graph from scratch using your own commonsense knowledge or the domain knowledge you want your AI to know about. But if you can extract knowledge from text, you can build much larger knowledge graphs much more quickly. Plus, you will need this algorithm to double-check any text generated by your language models. Knowledge extraction requires