Cosine similarity measures how similar two Vectors are by taking the cosine of the angle between them, producing a value between -1 and 1 that is invariant to magnitude. Borrowed from decades-old information-retrieval research, the metric has become the near-default way to compare Embeddings in AI systems. RAG pipelines, semantic search engines and recommendation systems all answer the question 'which document is closest to this query?' by computing cosine similarity. It is cheap to calculate — for normalised vectors it reduces to a dot product — which is why it scales comfortably to millions of vectors.
MEVZU N°124ISTANBULYEAR I — VOL. III
Glossary · Beginner · 1950
Cosine Similarity
A similarity measure based on the angle between two vectors, returning a value between -1 and 1.
- EN — English term
- Cosine Similarity
- TR — Turkish term
- Kosinüs Benzerliği