AI Terminology
/Intermediate
Vector Search
Definition
The overarching technology of querying data based on its meaning and context (semantic similarity) rather than exact lexical (keyword) matches.
Explain Like I'm New
If you search for 'Canine doctor', a keyword search looks for those exact words and fails. A Vector Search understands that 'Canine doctor' means exactly the same thing as 'Veterinarian for dogs' and successfully finds the document.
Real World Example
Google Image Search. You upload a picture of a red chair. The AI converts your picture into a vector, runs a Vector Search across billions of other images, and shows you visually similar red chairs.
Common Use Cases
- •Semantic search
- •Image matching
- •Product recommendations
Interview Questions
basic
- Does Vector Search require the query words to exactly match the document words?
intermediate
- Why do we convert data into Vectors before searching?