知识卡片
歧义词不能仅按负面分数处理
mean可能是数学平均,也可能形容苛刻;没有足够上下文时,意图分数不能替系统确定含义。可优先处理明确答案并对不确定部分澄清,避免因某标签较高就把正常学习语句当侮辱。
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
This multi-label classifier is built from multiple independent binary classifiers, one for each possible label (tag). Having independent tags for each possible intent gives you fine-grained control over your chatbot’s decisions and responses. For example, you could ignore insults in the messages and extract the numerical answer if the model predicts a label such as “answer” or “numerical expression” with high confidence. It may be
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:摘录中关于歧义词不能仅按负面分数处理的定义、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:This multi-label classifier is built from multiple independent binary classifiers, one for each possible label (tag). Having independent tags for each possible intent gives you fine-grained control over your chatbot’s decisions and responses. For example, you could ignore insults in the messages and extract the numerical answer if the model predicts a label such as “answer” or “numerical expression” with high confidence. It may be