Multi-match query in Elasticsearch - Step-by-Step Execution

Concept Flow - Multi-match query

Start Query

↓

Specify Multi-match Query

↓

Provide Search Text

↓

Define Fields to Search

↓

Elasticsearch Searches All Fields

↓

Combine Scores from Fields

↓

Return Best Matches

↓

End

The multi-match query takes a search text and looks for it across multiple fields, combining scores to find the best matches.

Execution Sample

Elasticsearch

{
  "query": {
    "multi_match": {
      "query": "apple",
      "fields": ["title", "description"],
      "type": "most_fields"
    }
  }
}

This query searches for the word 'apple' in both the 'title' and 'description' fields.

Execution Table

Step	Action	Input	Fields Searched	Resulting Score	Notes
1	Start multi-match query	query='apple'	title, description	N/A	Query initialized
2	Search 'apple' in 'title'	apple	title	0.8	Found 'apple' in title with score 0.8
3	Search 'apple' in 'description'	apple	description	0.5	Found 'apple' in description with score 0.5
4	Combine scores	0.8 + 0.5	title, description	1.3	Scores combined for final relevance
5	Return document with combined score	score=1.3	title, description	1.3	Document returned as a match
6	End	No more docs	N/A	N/A	Query execution complete

💡 All specified fields searched and scores combined; query ends after returning matches.

Variable Tracker

Variable	Start	After Step 2	After Step 3	After Step 4	Final
query	N/A	apple	apple	apple	apple
fields	N/A	[title, description]	[title, description]	[title, description]	[title, description]
score_title	0	0.8	0.8	0.8	0.8
score_description	0	0	0.5	0.5	0.5
combined_score	0	0	0	1.3	1.3

Key Moments - 2 Insights

Why does the multi-match query combine scores from multiple fields?

Because it searches the same query text in each specified field separately and then adds their relevance scores to find the best overall match, as shown in steps 2, 3, and 4 of the execution table.

What happens if the query text is found in only one field?

Only that field contributes a score, and the combined score equals that single field's score. This is implied in the execution table where scores are added only if found.

Visual Quiz - 3 Questions

Look at the execution table, what is the combined score after searching both fields?

A 0.5

B 0.8

C 1.3

D 2.0

💡 Hint

Check the 'Combine scores' row (Step 4) in the execution table.

At which step does Elasticsearch search the 'description' field?

A Step 3

B Step 2

C Step 4

D Step 5

💡 Hint

Look at the 'Action' column for searching fields in the execution table.

If the query text was not found in 'title', how would the combined score change?

A It would be 1.3

B It would be 0.5

C It would be 0

D It would be 0.8

💡 Hint

Refer to the variable_tracker for scores after each step.

Concept Snapshot

Multi-match query syntax:
{
  "query": {
    "multi_match": {
      "query": "text",
      "fields": ["field1", "field2"],
      "type": "most_fields"
    }
  }
}

Behavior: Searches given text across multiple fields, combines relevance scores, and returns best matches.
Key rule: Scores from each field add up to determine overall relevance.

Full Transcript

The multi-match query in Elasticsearch lets you search for a word or phrase across several fields at once. You start by specifying the query text and the fields you want to search. Elasticsearch then looks for the text in each field separately and gives each match a score. These scores are added together to find the best overall matches. For example, if the word 'apple' is found in the 'title' field with a score of 0.8 and in the 'description' field with a score of 0.5, the combined score is 1.3. The documents with the highest combined scores are returned as the best matches. This way, you can search multiple parts of your data with one query and get results ranked by how well they match across all fields.