Let say I have thousands of articles indexed.
When a user visit one of these articles at my website, I wanna show related articles.
Right now, I’m first getting the vector from pinecone by the ID of the article that the user is visiting, then I’m querying pinecone using this vector for similar articles.
This seems like a lot of overhead though, by selecting the vector and then using it for the next query.
is there a way to query this directly using the ID of existing document?
I’m thinking something like this
index.query(
id=“123”,
top_k=3
)