Vector Search

Vector search is a search method which returns the result which are close to your search query in multi dimensional space.

The basic difference between Full Text search and Vector Search is that it in FTS it returns only data based on the text-matches, whereas in Vector Search you get the result which are close to the vectors in multi-dimensional field.

Concepts

Vector

A vector is a 1-D numerical array that represents data across multiple dimensions. Vectors can encapsulate various types of data, including text, images, audio, and unstructured data. Semantic similarity between vectors is assessed by calculating the distance between them.

Vector Embedding

Vector Embedding also known as vectorization is the process of converting the data into vectors.

Embedding Model

The LLM Models which convert normal data to Vector Embedding, for example OpenAI GPT model, AWS Bedrock, Google Gemini AI, etc.

Last updated