Full Text Search Force Find All Tokens

Elastic Stack Elasticsearch
1

I try to implement full text search.
I have a question related to this subject.
I want to at search time, when i search a text, for example "Hello Saeed", return result only when all tokens of that text exist in a message.
this means in "Hello Saeed" that contains "Hello" and "Saeed" tokens just must be return messages that exist both of the words not one of them.

Hello I am Saeed ---> must be return
Hello I am Mike ---> must not be return (because at search time i searched "Hello Saeed" and 'Saeed' not exist in this sentence)

but for search "Hello" both of the previous sentences are approve.

finally my question is, how can i force text analyzer at search time to find all tokens are exist in search text?

2

Have a look at the minimum_should_match parameter.

3

My query is

GET /message/_search
{
  "from": 0,
  "size": 10,
  "sort": {
    "@timestamp": "desc"
  },
  "query": {
    "bool": {
      "must": [
        {
          "terms": {
            "roomId": [
              1,
              2
            ]
          }
        },
        {
          "match": {
            "message": "Hello Saeed"
          }
        },
        {
          "match": {
            "type": 0
          }
        },
        {
          "range": {
            "@timestamp": {
              "gte": 1000
            }
          }
        }
      ]
    }
  }
}

what i know, minimum_should_match use for define minimum number should be match in list of match case.
but in my question i want to force tokens existence that exist in a text message in ONE match item not multiple. so seems to minimum_should_match is not useful here.

4

It is an option for your match query.

5

operator in match query is exactly my answer.

"operator = and" force match query to find all tokens that created previously by tokenizer in message analyzer

thanks

6

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.