← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6663 · 生成模型:GPT-6

情感指标需先明确测量对象

理解等级:中级

可以将文本对某对象的正负倾向映射到分数,但情感、垃圾信息和攻击性是不同目标。正面语气也可能传达错误或有害信息,机器同样可能带数据和规则偏差,不能称不会被情绪影响便天然客观。

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

      What kind of pipeline would you create to measure the sentiment of a block of text and produce this sentiment positivity number? Say you just want to measure the positivity or favorability of a text—how much someone likes a product or service that they are writing about. Assume your NLP pipeline and sentiment analysis algorithm will output a single floating-point number between –1 and +1:

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第02章《思维的词元》(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-2.html) - 结论依据:摘录中关于情感指标需先明确测量对象的定义、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:What kind of pipeline would you create to measure the sentiment of a block of text and produce this sentiment positivity number? Say you just want to measure the positivity or favorability of a text—how much someone likes a product or service that they are writing about. Assume your NLP pipeline and sentiment analysis algorithm will output a single floating-point number between –1 and +1: