KStem token filter

Provides KStem-based stemming for the English language. The kstem filter combines algorithmic stemming with a built-in dictionary.

The kstem filter tends to stem less aggressively than other English stemmer filters, such as the porter_stem filter.

The kstem filter is equivalent to the stemmer filter’s light_english variant.

This filter uses Lucene’s KStemFilter.

Example

The following analyze API request uses the kstem filter to stem the foxes jumping quickly to the fox jump quick:

GET /_analyze
{
  "tokenizer": "standard",
  "filter": [ "kstem" ],
  "text": "the foxes jumping quickly"
}

The filter produces the following tokens:

[ the, fox, jump, quick ]

Add to an analyzer

The following create index API request uses the kstem filter to configure a new custom analyzer.

To work properly, the kstem filter requires lowercase tokens. To ensure tokens are lowercased, add the lowercase filter before the kstem filter in the analyzer configuration.

PUT /my_index
{
  "settings": {
    "analysis": {
      "analyzer": {
        "my_analyzer": {
          "tokenizer": "whitespace",
          "filter": [
            "lowercase",
            "kstem"
          ]
        }
      }
    }
  }
}