boost exact phrase search results ranking in elasticsearch

Elasticsearch use the DSL format to create query.

One easy search is to use multi_match by passing the query key word, and give the fields to search for.

Here is an example:

from elasticsearch import Elasticsearch
es = Elasticsearch()
indexname = 'myindex'

keyword = 'test'
   dsl={
       "query": {
           "multi_match" : {
                   "query":  keyword,
                   "fields": [ "content"]
               }

           }
 
   }


   
   dsize=10
   result_r = es.search(index=indexname, body=dsl,size=dsize)

The problem with the above query is that:if your query keyword is a phrase, you might find many results that have the exact match are ranked lower.

To solve this problem, you might want to try this new dsl format:

dsl={
   
    "query": {
       "bool": {
         "must": [
           {
             "multi_match": {
               "query": keyword,
               "fields": [
                  "content1","content2"
               ]
             }
           }
         ],
         "should": [
           {
             "multi_match": {
               "query": keyword,
               "fields": [
                  "content1","content2"
               ],
               "type": "phrase",
               "boost": 10
             }
           },
           {
             "multi_match": {
               "query": keyword,
               "fields": [
               "content1","content2"
               ],
               "operator": "and",
               "boost": 4
             }
           }
         ]
       }
     }
           
   }

one thing to explain in the above the query, is the ‘operator’; According to the official elasticsearch webpage:

operator and minimum_should_match
The best_fields and most_fields types are field-centric — they generate a match query per field. This means that the operator and minimum_should_match parameters are applied to each field individually, which is probably not what you want.

Take this query for example:

GET /_search
{
  "query": {
    "multi_match" : {
      "query":      "Will Smith",
      "type":       "best_fields",
      "fields":     [ "first_name", "last_name" ],
      "operator":   "and" 
    }
  }
}

This query is executed as:

(+first_name:will +first_name:smith)
| (+last_name:will +last_name:smith)

In other words, all terms must be present in a single field for a document to match.