知识卡片
一条消息可以同时有多个意图
学生可能同时给出数值答案和不满反馈,多标签输出能保留这两部分。每类可独立计分,但如何执行仍由业务优先级与语境决定;多标签模型不必由完全独立的多个网络实现。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
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 most important piece of NLU your chatbot will need to do is recognize your user’s intent. Users are often ambiguous and express many different sentiments and goals in one message, so you will need a multi-label classifier to capture all of their intents. If you are building a math tutor bot, you will need to be able to extract and recognize numerical (math) expressions.
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:摘录中关于一条消息可以同时有多个意图的定义、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 most important piece of NLU your chatbot will need to do is recognize your user’s intent. Users are often ambiguous and express many different sentiments and goals in one message, so you will need a multi-label classifier to capture all of their intents. If you are building a math tutor bot, you will need to be able to extract and recognize numerical (math) expressions.