An implementation of BM25 variants surveyed in "Improvements to BM25 and Language Models Examined" by Trotman et al., 2014.
gokapiX is a ranking algorithm designed to implement multiple BM25 variants, improving upon traditional BM25 scoring for information retrieval. This implementation is inspired by rank_bm25.
Name Credits: gokapi
Algorithms implemented:
- Atire BM25
- BM25L
- BM25+
- BM25-Adpt
- BM25T
To install gokapiX, use:
go get github.com/niranjanorkat/gokapiX