知识卡片
语义检索仍在估计用户意图
嵌入可将不同拼写而常见于相似语境的表达放得较近,帮助扩展词面匹配。模型距离仍只是训练数据与表示下的相关性线索,不是直接读取用户真实意图;专名、数字和精确术语仍可能需要词项匹配共同支持。
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
- · 用词嵌入推理
原书出处 · 1 处
- · 查看原书依据
Then, information retrieval researchers discovered how to make latent semantic analysis (LSA) more effective—word embeddings. These new word embeddings (vectors) made it possible for search engines to directly match the “meaning” of your query to web pages, without having to guess your intent. The embeddings for your search terms provide a direct numerical representation of the intent of your search based on the average meaning
参考来源
- 位置:《Natural Language Processing in Action, Second Edition》第06章《用词嵌入推理》(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-6.html)
- 结论依据:摘录中关于语义检索仍在估计用户意图的定义、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:Then, information retrieval researchers discovered how to make latent semantic analysis (LSA) more effective—word embeddings. These new word embeddings (vectors) made it possible for search engines to directly match the “meaning” of your query to web pages, without having to guess your intent. The embeddings for your search terms provide a direct numerical representation of the intent of your search based on the average meaning