Temporal Word Analogies: Identifying Lexical Replacement with Diachronic Word Embeddings
Temporal Word Analogies: Identifying Lexical Replacement with Diachronic Word Embeddings
复制标题
时间词类比：识别历时词嵌入的词汇替换
DOI：
--
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
Terrence Szymanski
中科院分区：
文献类型：
--
作者：
Terrence Szymanski
本文介绍了时间词类比的概念：在不同的时间点占据相同语义空间的词对。词嵌入的一个众所周知的特性是，它们能够通过向量加法有效地模拟传统的词类比（“词w_1是词w_2，就像词w_3是词w_4”）。在这里，我展示了时间词类比（“在时间t_alpha的词w_1就像在时间t_alpha的词w_2”），eta”）可以有效地用历时词嵌入来建模，只要来自每个时间段的独立嵌入空间被适当地变换成公共向量空间。当应用于一个历时语料库的新闻文章，这种方法是能够识别时间词类比，如“罗纳德里根在1987年就像比尔克林顿在1997年”，或“随身听在1987年就像iPod在2007年”。
This paper introduces the concept of temporal word analogies: pairs of words which occupy the same semantic space at different points in time. One well-known property of word embeddings is that they are able to effectively model traditional word analogies (“word w_1 is to word w_2 as word w_3 is to word w_4”) through vector addition. Here, I show that temporal word analogies (“word w_1 at time t_alpha is like word w_2 at time t_eta”) can effectively be modeled with diachronic word embeddings, provided that the independent embedding spaces from each time period are appropriately transformed into a common vector space. When applied to a diachronic corpus of news articles, this method is able to identify temporal word analogies such as “Ronald Reagan in 1987 is like Bill Clinton in 1997”, or “Walkman in 1987 is like iPod in 2007”.