Loading timeline…
20132010s
Word2Vec
Highly scalable models to produce dense vector representations of language.
Why It Was Important
Developed by Tomas Mikolov at Google, Word2Vec learned to map words to vectors based on their context within billions of words. It spectacularly captured semantic relationships mathematically (e.g., Vector('King') - Vector('Man') + Vector('Woman') ≈ Vector('Queen')), forever altering natural language processing.
Who Invented It
Tomas Mikolov et al.
Information retrieval specialist at Google.
Applications
- Semantic Search
- NLP Pipelines
- Language Translation Embedding
Key Papers
- Efficient Estimation of Word Representations in Vector Space
Tomas Mikolov et al. · ICLR 2013 Workshop
Videos
Word Embedding and Word2Vec, Clearly Explained!!!
StatQuest with Josh Starmer
What are Word Embeddings?
IBM Technology