Autocomplete with edge n-gram in Elasticsearch - Mini Project: Build & Apply

Autocomplete with edge n-gram

📖 Scenario: You are building a search feature for a bookstore website. You want users to get suggestions as they type book titles.

🎯 Goal: Create an Elasticsearch index with edge n-gram analyzer to enable autocomplete suggestions on book titles.

📋 What You'll Learn

Create an index called books with edge n-gram analyzer

Add a title field using the edge n-gram analyzer

Index sample book documents with exact titles

Query the index to get autocomplete suggestions for a given prefix

💡 Why This Matters

🌍 Real World

Autocomplete helps users find books quickly by suggesting titles as they type, improving user experience on bookstore websites.

💼 Career

Understanding edge n-gram analyzers and autocomplete queries is useful for search engineers and backend developers working with Elasticsearch.

1

Create the books index with edge n-gram analyzer

Write a JSON request to create an Elasticsearch index called books with a custom analyzer named autocomplete that uses an edge_ngram tokenizer with min_gram 1 and max_gram 10. Define a title field of type text that uses this analyzer for indexing and a standard analyzer for searching.

Elasticsearch

{
  "settings": {
    "analysis": {
      "analyzer": {
        "autocomplete": {
          "tokenizer": "edge_ngram_tokenizer"
        }
      },
      "tokenizer": {
        "edge_ngram_tokenizer": {
          "type": "edge_ngram",
          "min_gram": 1,
          "max_gram": 10,
          "token_chars": ["letter", "digit"]
        }
      }
    }
  },
  "mappings": {
    "properties": {
      "title": {
        "type": "text",
        "analyzer": "autocomplete",
        "search_analyzer": "standard"
      }
    }
  }
}
# Your code here

Need a hint?

Define the settings.analysis.analyzer.autocomplete with the edge_ngram_tokenizer. Then map the title field to use this analyzer.

2

Index sample book documents

Index three book documents into the books index with the exact title values: "The Great Gatsby", "Gone with the Wind", and "Great Expectations".

Elasticsearch

{
  "settings": {
    "analysis": {
      "analyzer": {
        "autocomplete": {
          "tokenizer": "edge_ngram_tokenizer"
        }
      },
      "tokenizer": {
        "edge_ngram_tokenizer": {
          "type": "edge_ngram",
          "min_gram": 1,
          "max_gram": 10,
          "token_chars": ["letter", "digit"]
        }
      }
    }
  },
  "mappings": {
    "properties": {
      "title": {
        "type": "text",
        "analyzer": "autocomplete",
        "search_analyzer": "standard"
      }
    }
  }
}

POST /books/_doc/1
{
  "title": "The Great Gatsby"
}

POST /books/_doc/2
{
  "title": "Gone with the Wind"
}

POST /books/_doc/3
{
  "title": "Great Expectations"
}
# Your code here

Need a hint?

Use POST /books/_doc/<id> to add each book with the exact title.

3

Write the autocomplete query

Write a JSON query to search the books index for autocomplete suggestions on the title field using a match query with the prefix "Gre".

Elasticsearch

{
  "query": {
    "match": {
      "title": "Gre"
    }
  }
}
# Your code here

Need a hint?

Use a match query on the title field with the value "Gre" to get autocomplete results.

4

Display the autocomplete results

Print the titles of the books returned by the autocomplete query for the prefix "Gre". Assume the query result is stored in a variable called response and print each matching _source.title on a new line.

Elasticsearch

for hit in response["hits"]["hits"]:
    print(hit["_source"]["title"])
# Your code here

Need a hint?

Loop over response["hits"]["hits"] and print each _source.title.