TinkerThe search waits on the desk.
Chapter 4 of 7
Long pages and repeats
Rarity fixed the forklift search. Now someone searches “price”, and the winner is simply the page that says it most.
This chapter: Should a page that repeats a word, or a long page, win?
About this chapter
You will learn
- Each repeat adds less than the last
- Long pages are scored down
- BM25 does both
Words you’ll meet
- BM25
- The scoring most keyword search uses: rarity, repeats that help less each time, and a fair share for length.
- Saturation
- Each repeat of a word adding less to the score than the one before.
Step 1 of 5
Weighed by rarity, every repeat of a word adds the same again, and a long page has more room for repeats.
Search for “price”.