Multi words synonyms with tokens between the synonym's parts

[Olaph (Ian)]

Hi,

given, that I have following synonyms configuration:

"želvárium,aquaterrarium for turtles"
(želvárium is czech short-term for aquarium for turtles)

Is there a way that allowes me to match a document with following title when searching "želvárium"?

"pacific aquaterrarium 50x25x19cm for water turtles"

With the settings below, I can match only "pacific aquaterrarium for turtles" - there is a problem with the tokens "50x25x19cm" and "water" between "aquaterrarium" and "for turtles".

Setting for this example:

PUT /syntest
{
  "settings": {
    "index": {
      "analysis": {
        "analyzer": {
          "my_synonyms": {
            "tokenizer": "standard",
            "filter": [
                "my_synonym_graph"
            ]
          }
        },
        "filter": {
          "my_synonym_graph": {
            "type": "synonym_graph",
            "lenient": true,
            "synonyms": [
              "želvárium,aquaterrarium for turtles"
            ]
          }
        }
      }
    }
  },
  "mappings": {
    "properties": {
      "title": {
         "type": "text",
         "search_analyzer": "my_synonyms"
      }
    }
  }  
}

POST /syntest/_doc
{
  "title": "pacific aquaterrarium 50x25x19cm for water turtles"
}

GET /syntest/_search
{
  "query" : {
    "match": {
      "title": {
        "query": "želvárium",
        "analyzer": "my_synonyms"
      }
    }
  }
}

[Olaph (Ian)]

Looks like one possible solution is to use match_phrase with slop parameter.
In this case:

GET /syntest/_search
{
  "query" : {
    "match_phrase": {
      "title": {
        "query": "želvárium",
        "analyzer": "my_synonyms",
        "slop": 2
       }
    }
  }
}

would allow to ignore those two tokens [50x25x19cm] and [water] and return the desired document.
Not sure if this is the right way hot to solve it.