Word Embedding
/wɜːrd ɪmˈbɛdɪŋ/
Dense vector representations of words that capture semantic and syntactic relationships.
In italiano: Word EmbeddingWord embeddings map words to continuous vectors where semantically similar words are close in vector space. Learned from large corpora, they enable transfer learning in NLP. Popular methods: Word2Vec, GloVe, FastText.
Examples
- Word2Vec embeddings
- GloVe vectors
- Contextual embeddings from BERT