Vector search algorithm for hybrid search

General
1

I was reading the blog post: Unlocking the Power of Hybrid Search - A Deep Dive into Weaviate's Fusion Algorithms | Weaviate - Vector Database
In this, ranked and relative score fusion are explained. Maybe I have read over it, but it did not describe what algorithms were used for vector search. I did find bm25f for index search, but not vector search.

Anybody know what algorithm(s)/ technique(s) are used for vector search during the two mentioned hybrid searches?

2

Hi @Joey_Visbeen ! Welcome to our community!

You can choose between different options:

Here we have a blog on that (be aware: it’s quite a rabbit hole :joy: ):

This is how you can set that configuration using python client v4:

import weaviate.classes as wvc

client.collections.create(
    "Article",
    vector_index_config=wvc.config.Configure.VectorIndex.hnsw(
        distance_metric=wvc.config.VectorDistances.COSINE
    ),
)

Let me know if that helps :slight_smile:

1 Like
3

Thank you, that was the answer I was looking for. I don’t mind a rabbit hole! :grinning:

1 Like