Performing similarity search

Now that we have created an index that’s suitable to the vector length and type (FLOAT32 or FLOAT64), and we have chosen the right distance concerning the data represented by the vector embeddings (we care about the orientation of the vectors, or we are also interested in their magnitude), we can perform a comparison. Let’s consider a simple scenario based on the former example. We have three sentences in our database stored as Hashes, along with their vector embeddings:

• doc:1 stores This is a technical document, it describes the SID sound chip of the Commodore 64
• doc:2 stores The Little Prince is a short story by Antoine de Saint-Exupéry, the best known of his literary productions, published on April 6, 1943 in New York
• doc:3 stores Pasta alla carbonara is a characteristic dish of Lazio and more particularly of Rome, prepared with popular ingredients and with an intense flavour.

We will calculate the corresponding...