知识卡片
字符特征复用局部拼写片段
字符1到3元组使不同整词可共享局部片段,帮助从已见形式向新拼写迁移。共享片段并非必然同义或同标签;应在正确隔离的样本上检验,而非由三元组流行就断定它是所有文本任务的最佳范围。
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
- · 神经网络基础
原书出处 · 1 处
- · 查看原书依据
But how can you tokenize a single word, like a name, so that your model can generalize to completely original names that it’s never seen before? You can use the character n-grams within each word (or name) as your tokens, and you can set up a TfidfVectorizer to count characters and character n-grams rather than words. You can experiment with a wider or narrower ngram_range,
参考来源
- 位置:《Natural Language Processing in Action, Second Edition》第05章《神经网络基础》(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-5.html)
- 结论依据:摘录中关于字符特征复用局部拼写片段的定义、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:But how can you tokenize a single word, like a name, so that your model can generalize to completely original names that it’s never seen before? You can use the character n-grams within each word (or name) as your tokens, and you can set up a TfidfVectorizer to count characters and character n-grams rather than words. You can experiment with a wider or narrower ngram_range,