← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6784.b · 生成模型:GPT-6

循环要提供退出与帮助路径

理解等级:中级

重复练习或澄清可形成对话环,用户应能离开、求助或换任务。次数上限可以结合问题和上下文设置,而不是所有场景固定三次;无限自动环与有用户控制的多轮练习也应区别检查。

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

      In a decision tree, you are not allowed to jump back to a previous decision because this might create a perpetual recursive loop, where your algorithm never makes a decision. In a conversation graph, however, it is okay for people (and bots) to repeat themselves. The key to good conversation design is to make sure the user has a way to break out of any

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第12章《对话引擎与会话设计》(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-12.html) - 结论依据:摘录中关于循环要提供退出与帮助路径的定义、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:In a decision tree, you are not allowed to jump back to a previous decision because this might create a perpetual recursive loop, where your algorithm never makes a decision. In a conversation graph, however, it is okay for people (and bots) to repeat themselves. The key to good conversation design is to make sure the user has a way to break out of any