Skip to content
An Elasticsearch plugin for rescoring based on Redis keys
Branch: master
Clone or download

Latest commit

Fetching latest commit…
Cannot retrieve the latest commit at this time.


Type Name Latest commit message Commit time
Failed to load latest commit information.
gradle/wrapper Init Jan 22, 2020
src Fixing rescoring bug + protective coding + logging Feb 18, 2020
LICENSE.txt License and Notice files Jan 26, 2020
NOTICE.txt License and Notice files Jan 26, 2020 Update README Feb 20, 2020
build.gradle Configurable redis url Feb 13, 2020
docker-compose.yml Configurable redis url Feb 13, 2020 Init Jan 22, 2020
gradlew Init Jan 22, 2020
gradlew.bat Init Jan 22, 2020
settings.gradle Init Jan 22, 2020

Elasticsearch Redis Rescore Plugin

Ever wanted to use data from external systems to influence scoring in Elasticsearch? now you can.

The idea is simple: query a low-latency service per document during search and get a numeric value for that specific document, then use that number as a multiplier for influencing the document score, up or down.

This plugin uses Redis to rescore top ranking results in Elasticsearch. It's your job to make sure the query gets the basic filtering and scoring right. Then, you can use the plugin to rescore the top N results.

Rescoring with this plugin assumes Redis contains keys that can be correlated with the data in the documents, such that for every doc D there exists a value in a predefined field that also exists in Redis as a key whose value is numeric.

Documents that do not contain this field, or no value in this field, or that value does not exist in Redis as a key - are left untouched.

See an example below.


Follow the standard plugin installation instructions, with a zip version of the compiled plugin:


  "query": { "match_all":  {} },
  "rescore": {
      "key_field": "productId.keyword",
      "key_prefix": "mystore-"

In this example, we are expecting each hit to contain a field productId (of keyword type). The value of that field will be looked up in Redis as a key (for example, Redis key mystore-abc123 will be looked-up for a document with productId abc123; the mystore- key prefix is configurable in query time).

The value which will be found under that Redis key, if exists and of numeric type, will be multiplied by the current document score to produce a new score.

You can use 0 to demote results (e.g. mark as unavailable in stock), 1 to leave unchanged, or any other value to produce positive or negative boosts based on your business logic.

You can’t perform that action at this time.