Embedding
In plain terms
Turning words, images, or items into numeric vectors that capture meaning.
Going deeper
Embeddings map data into a space where similar things sit close together, letting machines compute similarity and feed neural networks. They're central to search, recommendations, and language models.
Under the hood
Learned vector spaces capture semantic relationships (king − man + woman ≈ queen); dimensionality and training objective shape what 'similarity' means.
Examples
word2vec, GloVe, OpenAI text embeddings, recommendation item vectors, vector DB search.
Think of it like...
Like placing every word on a giant map where synonyms cluster nearby: 'king' and 'queen' end up in the same neighborhood.