DP-800: Developing AI-Enabled Database Solutions

By Microsoft · Question Mode

QuestionQ35

Implement AI capabilities in database solutions

You have an Azure SQL database containing the following tables and columns.

Question Image

Embeddings in the NotesEmbeddings and DescriptionEmbeddings tables were generated from the Description and Notes values in the Articles table by using different chunk sizes.

You need to run approximate nearest neighbor (ANN) queries across both embedding tables while minimizing the effect of the different chunk sizes.

What should you use?

Select
Function:
Distance metric:
Explanation

VECTOR_SEARCH performs approximate nearest-neighbor retrieval over vector columns. Cosine distance is insensitive to vector magnitude because it measures the angle between embeddings, which reduces the impact that different text chunk sizes can have on embedding magnitude.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!