feat: Add blog draft for the Sparse Vectors based RecSys

[NirantK]

No description provided.

[agourlay]

compute user_sparse_vectors[row.user_id] only once and bind it to a variable :)

[mjang]

@NirantK , as I read through your work, I realized this sounded similar to https://qdrant.tech/documentation/tutorials/search-beginners/

IOW, I'm tempted to set this up as a new doc page, something like "Semantic search for advanced users". (Does that make sense? Does this include more advanced techniques?)

[NirantK]

@mjang This is not a beginner-friendly article or blog at all. This is a novel and experimental way to model recommendation systems in Qdrant from @generall. I'm inclined to believe that this is best treated as either an article or a blog?

[mjang]

@mjang This is not a beginner-friendly article or blog at all. This is a novel and experimental way to model recommendation systems in Qdrant from @generall. I'm inclined to believe that this is best treated as either an article or a blog?

Thanks for explaining, @NirantK . We'll proceed with this as an article, see what the reaction is. If it's positive, I might revisit the idea.