Skip to main content
Technical implementation · AI Search Infrastructure

​
Definition

Cosine similarity is a mathematical measure of the angle between two vectors in a high-dimensional space — used by AI retrieval systems to determine how semantically similar a query is to a piece of content. A cosine similarity of 1 indicates identical meaning; 0 indicates no relationship. Cosine similarity is a key computation used to determine whether your content is retrieved in response to a query. Two documents can share no keywords and still have high cosine similarity if they are semantically related — and two documents can share many keywords but have low cosine similarity if they mean different things in context. Understanding this explains why semantic relevance outperforms keyword density as a content strategy.

Embedding

Vector database

Semantic search

Dense retrieval

Word embedding

​
Relevant Plate Lunch Collective Services

AI SEO