知识卡片

普通读书笔记卡 · 6680.b · 生成模型:GPT-6

语义问答需复用完整变换链

理解等级:中级

训练时拟合TF–IDF与主题投影,新问题必须依次调用这两个已拟合变换,再在同一坐标比较文档。书中n_iterations和transformnp.asarray拼写无效;修正接口后仍需核对归一化、行映射、拒答与独立问题表现,不能用一条成功例证明健壮。

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 next step is to represent both the questions and our query as vectors. This is where we need to add just a few lines to make our representations semantic. Because our question dataset is small, we won’t need to apply LSH or any other indexing algorithm; we can just go through our questions one by one and choose the best match.

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第04章《从词数到语义分析》(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-4.html) - 结论依据:摘录中关于语义问答需复用完整变换链的定义、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 next step is to represent both the questions and our query as vectors. This is where we need to add just a few lines to make our representations semantic. Because our question dataset is small, we won’t need to apply LSH or any other indexing algorithm; we can just go through our questions one by one and choose the best match.