A few weeks ago, I delivered some sessions about vector search in Fabric Databases and Azure SQL. Let’s talk about some advanced features available in both of them.
Before analysing these advanced features, we need to understand some concepts about vectors and embeddings.
Embeddings Concepts
Creating embeddings is the process of creating vectors for documents or images, representing the content of these elements.
Each vector dimension has a specific meaning. The precise meaning is not easy identifiable, it depends a lot of the model used to create the embedding. Let’s consider that identifying the meaning of each dimension is not