Completion suggester in Elasticsearch - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is the purpose of the Completion suggester in Elasticsearch?

The Completion suggester helps provide fast autocomplete or search-as-you-type functionality by suggesting possible completions for a given prefix.

beginner

Which field type must be used in the mapping to enable the Completion suggester?

The field must be mapped as completion type to enable the Completion suggester functionality.

intermediate

How do you define a Completion suggester query in Elasticsearch?

You define it inside the _search request under the suggest section, specifying the field, prefix, and size of suggestions.

intermediate

What is the difference between prefix and regex in Completion suggester queries?

prefix matches suggestions starting with the given text, while regex allows pattern matching for more flexible suggestions.

advanced

Why is the Completion suggester faster than a regular prefix query?

Because it uses a specialized data structure called a FST (Finite State Transducer) optimized for prefix lookups, making suggestions very fast.

Which field type enables the Completion suggester in Elasticsearch?

A completion

B text

C keyword

D integer

Only the 'completion' field type supports the Completion suggester functionality.

Where do you specify the Completion suggester in an Elasticsearch search request?

A In the 'aggs' section

B Inside the 'query' section

C In the 'sort' section

D Under the 'suggest' section

The Completion suggester is defined inside the 'suggest' section of the search request.

What does the 'prefix' parameter do in a Completion suggester query?

A Matches suggestions containing the prefix anywhere

B Matches suggestions ending with the prefix

C Matches suggestions starting with the prefix

D Matches suggestions exactly equal to the prefix

'prefix' matches suggestions that start with the given text.

Which data structure does the Completion suggester use for fast lookups?

A Linked List

B Finite State Transducer (FST)

C Binary Search Tree

D Hash Table

The Completion suggester uses a Finite State Transducer (FST) for efficient prefix matching.

Can the Completion suggester handle fuzzy matching by default?

A No, it only supports exact prefix matching

B Yes, it supports fuzzy matching by default

C Only if 'fuzzy' parameter is set to true

D Only with regex patterns

The Completion suggester does not support fuzzy matching by default; it matches exact prefixes.

Explain how to set up and use the Completion suggester in Elasticsearch.

Think about mapping, indexing, and querying steps.

Describe why the Completion suggester is efficient for autocomplete compared to regular text queries.

Focus on data structure and performance.

Practice

1. What is the main purpose of the completion suggester in Elasticsearch?

easy

A. To update documents in bulk

B. To perform full-text search on documents

C. To aggregate data based on terms

D. To provide fast autocomplete suggestions as users type

Solution

1. Step 1: Understand the role of completion suggester

   The completion suggester is designed to quickly suggest words or phrases as a user types input.

2. Step 2: Differentiate from other Elasticsearch features

   Full-text search, aggregation, and bulk updates serve different purposes unrelated to autocomplete suggestions.

3. Final Answer:

   To provide fast autocomplete suggestions as users type -> Option D

4. Quick Check:

   Completion suggester = autocomplete suggestions [OK]

Hint: Completion suggester = fast autocomplete feature [OK]

Common Mistakes:

• Confusing completion suggester with full-text search
• Thinking it aggregates data
• Assuming it updates documents

2. Which of the following is the correct way to define a completion field in an Elasticsearch mapping?

easy

A. {
"properties": {
"suggest": { "type": "completion" }
}
}

B. {
"properties": {
"suggest": { "type": "text" }
}
}

C. {
"mappings": {
"suggest": { "type": "completion" }
}
}

D. {
"fields": {
"suggest": { "type": "keyword" }
}
}

Solution

1. Step 1: Identify correct mapping structure

   The completion field must be defined inside properties with type set to completion.

2. Step 2: Compare options

   { "properties": { "suggest": { "type": "completion" } } } correctly uses properties and type: completion. { "properties": { "suggest": { "type": "text" } } } uses text type, which is incorrect. { "mappings": { "suggest": { "type": "completion" } } } wrongly places suggest directly under mappings. { "fields": { "suggest": { "type": "keyword" } } } uses fields and keyword type, which is incorrect.

3. Final Answer:

   { "properties": { "suggest": { "type": "completion" } } } -> Option A

4. Quick Check:

   Completion field in properties with type completion = { "properties": { "suggest": { "type": "completion" } } } [OK]

Hint: Completion field must be under properties with type 'completion' [OK]

Common Mistakes:

• Placing completion field outside properties
• Using text or keyword type instead of completion
• Misnaming mapping keys

3. Given the following suggest query, what will be the expected output if the index contains documents with suggest fields 'apple', 'applet', and 'application'?

{
  "suggest": {
    "fruit-suggest": {
      "prefix": "app",
      "completion": {
        "field": "suggest"
      }
    }
  }
}

medium

A. ["apple", "applet", "application"]

B. ["app", "apple", "applet"]

C. ["application"]

D. []

Solution

1. Step 1: Understand the prefix and completion suggester

   The prefix 'app' matches all suggest entries starting with 'app', which includes 'apple', 'applet', and 'application'.

2. Step 2: Predict the output list

   The completion suggester returns all matching suggestions starting with the prefix, so all three words will be suggested.

3. Final Answer:

   ["apple", "applet", "application"] -> Option A

4. Quick Check:

   Prefix 'app' matches all three suggestions [OK]

Hint: Prefix matches all suggestions starting with it [OK]

Common Mistakes:

• Expecting partial matches inside words
• Thinking prefix must be exact full word
• Assuming empty or partial results

4. Identify the error in this suggest query:

{
  "suggest": {
    "my-suggest": {
      "prefix": "ban",
      "completion": {
        "field": "suggestion"
      }
    }
  }
}

Assuming the mapping defines the completion field as suggest (not suggestion).

medium

A. The prefix value 'ban' is invalid

B. The suggest name 'my-suggest' is not allowed

C. The field name in completion should be 'suggest', not 'suggestion'

D. The suggest query must be inside 'query' key

Solution

1. Step 1: Check field name in completion suggester

   The completion suggester must reference the correct field name defined in the mapping. Here, the mapping uses 'suggest', but the query uses 'suggestion'.

2. Step 2: Validate other parts of the query

   The prefix 'ban' is valid, suggest name 'my-suggest' is allowed, and suggest queries do not require a 'query' key.

3. Final Answer:

   The field name in completion should be 'suggest', not 'suggestion' -> Option C

4. Quick Check:

   Field name mismatch causes error [OK]

Hint: Field name in query must match mapping exactly [OK]

Common Mistakes:

• Using wrong field name in completion query
• Thinking prefix value is invalid
• Expecting suggest query inside 'query' key

5. You want to implement a completion suggester that suggests city names but excludes any suggestions that start with the letter 'x'. Which approach correctly applies this filter in the suggest query?

hard

A. {
"suggest": {
"city-suggest": {
"prefix": "",
"completion": {
"field": "city_suggest",
"skip_duplicates": true
}
}
}
}

B. {
"suggest": {
"city-suggest": {
"prefix": "",
"completion": {
"field": "city_suggest",
"contexts": {
"first_letter": ["a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "y", "z"]
}
}
}
}
}

C. {
"suggest": {
"city-suggest": {
"prefix": "",
"completion": {
"field": "city_suggest",
"contexts": {
"first_letter": ["!x"]
}
}
}
}
}

D. {
"suggest": {
"city-suggest": {
"prefix": "",
"completion": {
"field": "city_suggest",
"filter": {
"term": { "first_letter": "x" }
}
}
}
}
}

Solution

1. Step 1: Understand filtering with contexts in completion suggester

   Completion suggester supports filtering using contexts. To exclude suggestions starting with 'x', you include all allowed first letters except 'x' in the context filter.

2. Step 2: Analyze each option

   { "suggest": { "city-suggest": { "prefix": "", "completion": { "field": "city_suggest", "skip_duplicates": true } } } } has no filtering. { "suggest": { "city-suggest": { "prefix": "", "completion": { "field": "city_suggest", "contexts": { "first_letter": ["!x"] } } } } } tries to exclude 'x' with '!x' which is invalid syntax. { "suggest": { "city-suggest": { "prefix": "", "completion": { "field": "city_suggest", "contexts": { "first_letter": ["a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "y", "z"] } } } } } explicitly lists all letters except 'x' in contexts, correctly filtering out 'x'. { "suggest": { "city-suggest": { "prefix": "", "completion": { "field": "city_suggest", "filter": { "term": { "first_letter": "x" } } } } } } uses a 'filter' key which is not supported in completion suggester.

3. Final Answer:

   { "suggest": { "city-suggest": { "prefix": "", "completion": { "field": "city_suggest", "contexts": { "first_letter": ["a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "y", "z"] } } } } } -> Option B

4. Quick Check:

   Use contexts with allowed letters excluding 'x' [OK]

Hint: Use contexts with allowed letters, exclude 'x' by omission [OK]

Common Mistakes:

• Using invalid '!x' syntax in contexts
• Trying to use 'filter' inside completion suggester
• Not applying context filtering at all