Synonym token filter

edit

The synonym token filter allows to easily handle synonyms during the analysis process.

Define synonyms sets

edit

Synonyms in a synonyms set are defined using synonym rules. Each synonym rule contains words that are synonyms.

You can use two formats to define synonym rules: Solr and WordNet.

Solr format
edit

This format uses two different definitions:

  • Equivalent synonyms: Define groups of words that are equivalent. Words are separated by commas. Example:

    ipod, i-pod, i pod
    computer, pc, laptop
  • Explicit synonyms: Matches a group of words to other words. Words on the left hand side of the rule definition are expanded into all the possibilities described on the right hand side. Example:

    personal computer => pc
    sea biscuit, sea biscit => seabiscuit
WordNet format
edit

WordNet defines synonyms sets spanning multiple lines. Each line contains the following information:

  • Synonyms set numeric identifier
  • Ordinal of the synonym in the synonyms set
  • Synonym word
  • Word type identifier: Noun (n), verb (v), adjective (a) or adverb (b).
  • Depth of the word in the synonym net

The following example defines a synonym set for the words "come", "advance" and "approach":

s(100000002,1,'come',v,1,0).
s(100000002,2,'advance',v,1,0).
s(100000002,3,'approach',v,1,0).""";

Configure synonyms sets

edit

Synonyms can be configured using the synonyms API, a synonyms file, or directly inlined in the token filter configuration. See store your synonyms set for more details on each option.

Use synonyms_set configuration option to provide a synonym set created via Synonyms Management APIs:

  "filter": {
    "synonyms_filter": {
      "type": "synonym",
      "synonyms_set": "my-synonym-set",
      "updateable": true
    }
  }

Synonyms sets must exist before they can be added to indices. If an index is created referencing a nonexistent synonyms set, the index will remain in a partially created and inoperable state. The only way to recover from this scenario is to ensure the synonyms set exists then either delete and re-create the index, or close and re-open the index.

Use synonyms_path to provide a synonym file :

  "filter": {
    "synonyms_filter": {
      "type": "synonym",
      "synonyms_path": "analysis/synonym-set.txt"
    }
  }

The above configures a synonym filter, with a path of analysis/synonym-set.txt (relative to the config location).

Use synonyms to define inline synonyms:

  "filter": {
    "synonyms_filter": {
      "type": "synonym",
      "synonyms": ["pc => personal computer", "computer, pc, laptop"]
    }
  }

Additional settings are:

  • updateable (defaults to false). If true allows reloading search analyzers to pick up changes to synonym files. Only to be used for search analyzers.
  • expand (defaults to true). Expands definitions for equivalent synonym rules. See expand equivalent synonyms.
  • lenient (defaults to the value of the updateable setting). If true ignores errors while parsing the synonym rules. It is important to note that only those synonym rules which cannot get parsed are ignored. See synonyms and stop token filters for an example of lenient behaviour for invalid synonym rules.
expand equivalent synonym rules
edit

The expand parameter controls whether to expand equivalent synonym rules. Consider a synonym defined like:

foo, bar, baz

Using expand: true, the synonym rule would be expanded into:

foo => foo
foo => bar
foo => baz
bar => foo
bar => bar
bar => baz
baz => foo
baz => bar
baz => baz

When expand is set to false, the synonym rule is not expanded and the first synonym is treated as the canonical representation. The synonym would be equivalent to:

foo => foo
bar => foo
baz => foo

The expand parameter does not affect explicit synonym rules, like foo, bar => baz.

tokenizer and ignore_case are deprecated
edit

The tokenizer parameter controls the tokenizers that will be used to tokenize the synonym, this parameter is for backwards compatibility for indices that created before 6.0. The ignore_case parameter works with tokenizer parameter only.

Configure analyzers with synonym token filters

edit

To apply synonyms, you will need to include a synonym token filters into an analyzer:

      "analyzer": {
        "my_analyzer": {
          "type": "custom",
          "tokenizer": "standard",
          "filter": ["stemmer", "synonym"]
        }
      }
Token filters ordering
edit

Order is important for your token filters. Text will be processed first through filters preceding the synonym filter before being processed by the synonym filter.

Elasticsearch will also use the token filters preceding the synonym filter in a tokenizer chain to parse the entries in a synonym file or synonym set. In the above example, the synonyms token filter is placed after a stemmer. The stemmer will also be applied to the synonym entries.

Because entries in the synonym map cannot have stacked positions, some token filters may cause issues here. Token filters that produce multiple versions of a token may choose which version of the token to emit when parsing synonyms. For example, asciifolding will only produce the folded version of the token. Others, like multiplexer, word_delimiter_graph or ngram will throw an error.

If you need to build analyzers that include both multi-token filters and synonym filters, consider using the multiplexer filter, with the multi-token filters in one branch and the synonym filter in the other.

Synonyms and stop token filters
edit

Synonyms and stop token filters interact with each other in the following ways:

Stop token filter before synonym token filter
edit

Stop words will be removed from the synonym rule definition. This can can cause errors on the synonym rule.

If lenient is set to false, invalid synonym rules can cause errors when applying analyzer changes. For reloadable analyzers, this prevents reloading and applying changes. You must correct errors in the synonym rules and reload the analyzer.

When lenient is set to false, an index with invalid synonym rules cannot be reopened, making it inoperable when:

  • A node containing the index starts
  • The index is opened from a closed state
  • A node restart occurs (which reopens the node assigned shards)

For explicit synonym rules like foo, bar => baz with a stop filter that removes bar:

  • If lenient is set to false, an error will be raised as bar would be removed from the left hand side of the synonym rule.
  • If lenient is set to true, the rule foo => baz will be added and bar => baz will be ignored.

If the stop filter removed baz instead:

  • If lenient is set to false, an error will be raised as baz would be removed from the right hand side of the synonym rule.
  • If lenient is set to true, the synonym will have no effect as the target word is removed.

For equivalent synonym rules like foo, bar, baz and expand: true, with a stop filter that removes `bar:

  • If lenient is set to false, an error will be raised as bar would be removed from the synonym rule.
  • If lenient is set to true, the synonyms added would be equivalent to the following synonym rules, which do not contain the removed word:
foo => foo
foo => baz
baz => foo
baz => baz
Stop token filter after synonym token filter
edit

The stop filter will remove the terms from the resulting synonym expansion.

For example, a synonym rule like foo, bar => baz and a stop filter that removes baz will get no matches for foo or bar, as both would get expanded to baz which is removed by the stop filter.

If the stop filter removed foo instead, then searching for foo would get expanded to baz, which is not removed by the stop filter thus potentially providing matches for baz.