本頁只刊出中文翻譯與中文說明；英文原文請見下方原文連結。
原文連結
論文資訊
- 類型：已發表論文
- 日期：2020-02-04
摘要
How does meaning vary across the world’s 語言s? Some claim word meanings are universal, and empirical research has shown cross-linguistic similarity in associations between concrete natural objects.1 Other research exploring specific domains such as color2 and kinship3 has demonstrated variability in the organization of word meanings by 語言, culture and environment. To what extent word meanings 文化ly vary or universally agree in different 語言s remains an unresolved empirical puzzle, but the 湧現 of powerful word embedding models4,5 enable us to take a systems-level approach that directly compares across semantic domains.6, 7 Here we show that meanings across 語言s are characterized by similarity within semantic domains and variability across them, using models trained on both (1) large corpora of n
※ 此為已發表論文，全文需透過期刊付費取得