We benchmarked Elasticsearch 8.14–8.18 as part of an upgrade we helped a customer with. We uncovered a hidden performance regression in nested indices and will share how we discovered this, along with the exact steps needed to fix it.
Lucene
In Lucene-based search engines like OpenSearch and Solr, keyword aggregations ignore duplicate values that occur within a multi-valued field. We built an OpenSearch plugin to overcome this limitation.
This post will show you precisely how BM25 builds upon TF-IDF, what its parameters do, and why it is so effective.
KMW hosted the Boston Elasticsearch meetup group, and our own Rudi Seitz gave an in-depth talk about how relevancy works in a search engine. In this post learn about when he covered, including: how scoring is computed, query structure and analysis chains, and approaches to tuning relevancy.