Database / Milvus Vector database Interview questions
What is an embedding vector, in the context of Milvus?
An embedding vector is a fixed-length array of floating-point numbers produced by a machine learning model to represent a piece of data, text, an image, audio, or something else, in a way that captures its semantic meaning geometrically: items with similar meaning end up as vectors that are close together in that high-dimensional space, according to some distance metric.
# conceptual example: two related sentences produce nearby vectors embed("a happy dog running in a park") # -> [0.12, -0.05, 0.88, ...] embed("a joyful puppy playing outdoors") # -> [0.14, -0.03, 0.85, ...] (close to the first) embed("quarterly tax filing deadline") # -> [-0.61, 0.42, 0.02, ...] (far from both)
Milvus itself doesn't generate embeddings; that's the job of an external embedding model (like an OpenAI, Cohere, or open-source sentence-transformer model). Milvus's role starts once those vectors exist: storing them alongside any related metadata, indexing them for efficient search, and finding the nearest vectors to a given query vector at search time.
More Related questions...
