Work

Embedding Learning Through Multilingual Concept Induction

Year: 2018
Type: preprint
Abstract: We present a new method for estimating vector space representations of words: embedding learning by concept induction. We test this method on a highly parallel corpus and learn semantic representations of words in 1259 different languages in a single common space. An extensive experimental evaluation on crosslingual word similarity and sentiment analysis indicates that concept-based multilingual embedding learning performs better than previous approaches.
Language: English
Open Access status: green
OpenAlex ID: W2953242939
Cites: 45
Cited by: 7
Related to: 9