Skip to content
BENIBIX Learn

Reciprocal Rank Fusion vs Reranking

People think reciprocal rank fusion is a kind of reranking, and it is not. Fusion merges two or more ranked lists using only each result's place in each list, so it is fast and reads no text. A reranker is a model that reads the question with each result and scores it again. Fusion joins the lists. The reranker then puts the short list in a better order.

AI🔎 Retrieval & Search

Reciprocal Rank Fusion (RRF)

A way to join ranked lists. Each place earns points.

Reciprocal rank fusion, or RRF, is a way to merge two or more ranked lists into one. Each list gives a result 1 divided by 60 plus its place. So first place earns 1/61, and tenth place earns 1/70. The points from every list are added, and the top total comes first. It looks only at places, not scores, so lists with different kinds of score can be merged. It was published in 2009.

Example
In Lucy's shop search, one page is 2nd in the keyword list and 5th in the meaning list. It earns 1/62 plus 1/65, which beats a page that is 1st in only one list.
Why it matters
It is the usual way hybrid search joins its two lists. It needs no tuning, and the two scores never have to be on the same scale.
Full entry
AI🔎 Retrieval & Search

Reranking

A second, slow look at the top hits. It puts them in a better order.

Reranking is a second pass that puts search results in a better order. A fast search first pulls, say, the top 50 candidates. Then a reranker model reads the question together with each candidate and gives it a new score. That is slower per text but much more accurate, so it is only run on the short list. The best few after reranking are the ones passed on to the user or the model.

Example
Jack's search pulls 50 help articles in 20 milliseconds. A reranker reads each with the question and moves the one true answer from 14th place to 1st.
Why it matters
It often gives the biggest jump in search quality for the least work.
Full entry

In one line

Reciprocal Rank Fusion (RRF)
A way to join ranked lists. Each place earns points.
Reranking
A second, slow look at the top hits. It puts them in a better order.