这里会显示出您选择的修订版和当前版本之间的差别。
两侧同时换到之前的修订记录 前一修订版 后一修订版 | 前一修订版 | ||
分布式表示 [2019/09/30 10:39] admin |
分布式表示 [2020/02/01 21:41] (当前版本) |
||
---|---|---|---|
行 1: | 行 1: | ||
====== 单词级别(word-level) ====== | ====== 单词级别(word-level) ====== | ||
+ | 词典:apple,car,cat.... | ||
===== 优势 ===== | ===== 优势 ===== | ||
单词有语义信息 | 单词有语义信息 | ||
行 11: | 行 12: | ||
====== 字符级别(character-level) ====== | ====== 字符级别(character-level) ====== | ||
+ | a-z,A-Z,个别符号 | ||
===== 缺点 ===== | ===== 缺点 ===== | ||
没有语义信息,需要神经网络学习语义信息 | 没有语义信息,需要神经网络学习语义信息 | ||
行 17: | 行 19: | ||
====== 单词级别和字符级别相互结合 ====== | ====== 单词级别和字符级别相互结合 ====== | ||
+ | ====== 参考文章 ====== | ||
+ | * [[https://www.lighttag.io/blog/character-level-NLP/|Character Level NLP:强烈推荐]] | ||
+ | * Character-level Convolutional Network for Text Classification Applied to Chinese Corpus | ||
+ | * Combining Word-Level and Character-Level Representations forRelation Classification of Informal Text | ||
+ | * Named Entity Recognition with Bidirectional LSTM-CNNs | ||