Searching Data in Elasticsearch

So far, you've learned how to:

• Store data in Elasticsearch.
• Organize data using indices and documents.
• Perform CRUD (Create, Read, Update, Delete) operations.

Now comes the most exciting part searching.

Searching is the primary reason Elasticsearch exists. While traditional databases can search data, Elasticsearch is specifically designed to search through massive amounts of information quickly and accurately.

Whether you're searching for a product on Amazon, a movie on Netflix, or an article on a news website, Elasticsearch helps deliver relevant results in milliseconds.

By the end of this lesson, you'll understand:

• How Elasticsearch performs searches.
• What a query is.
• Different types of search queries.
• How Elasticsearch decides which results are most relevant.
• The difference between searching and filtering.

A search query is simply a request asking Elasticsearch to find specific information.

Whenever you type something into a search box, you're creating a search query.

For example:

• Running Shoes
• Bluetooth Speaker
• Python Programming Book

These words are sent to Elasticsearch, which then searches through its indices to find matching documents.

Think of a search query as asking a librarian:

"Can you find books about space exploration?"

The librarian searches the library catalog instead of checking every book individually.

Elasticsearch works in a similar way.

Let's understand what happens behind the scenes.

Suppose a customer searches for:

Wireless Mouse

Elasticsearch follows these steps:

User Types Search

│

▼

Receive Search Query

│

▼

Analyze the Words

│

▼

Search the Index

│

▼

Find Matching Documents

│

▼

Rank Results by Relevance

│

▼

Display Results

Notice that Elasticsearch searches the index, not every document.

This is why searches are extremely fast.

Imagine searching an online bookstore for:

Python Programming

There might be hundreds of books containing the word Python.

Should every book appear in random order?

No.

Instead, Elasticsearch calculates which books are most relevant to your search.

For example:

The books that best match your search appear first.

This ranking process is one of Elasticsearch's biggest strength.

One of Elasticsearch's most powerful features is Full-Text Search.

Instead of looking for an exact word, Elasticsearch tries to understand what you're searching for.

Suppose a product is stored as:

Wireless Bluetooth Headphones

A customer searches:

Bluetooth Headset

Although the words aren't exactly the same, Elasticsearch may still return the correct product because it analyzes the text intelligently.

This creates a much better search experience than simple keyword matching.

The most common search type in Elasticsearch is the Match Query.

A Match Query searches analyzed text fields.

Imagine your products index contains these products:

A user searches:

Gaming

Elasticsearch returns:

• Gaming Laptop
• Gaming Keyboard

because both documents contain the searched word.

This type of search is flexible and works well for normal text searches.

A Term Query is different.

It looks for an exact value.

Imagine a field called:

Brand

Possible values:

• Dell
• HP
• Lenovo

If you search for:

Dell

Elasticsearch returns only documents where the brand is exactly Dell.

Term queries are commonly used with fields such as:

• Product IDs
• Brand names
• Status values
• Country codes

Unlike Match Queries, Term Queries do not analyze the text before searching.

Let's compare them.

Suppose a document contains:

Apple iPhone 16 Pro

Match Query

Searching:

Apple

or

iPhone

can successfully find the document because Elasticsearch analyzes the text.

Term Query

Searching:

Apple

will only work if the field stores Apple as an exact value.

If the field contains the complete phrase:

Apple iPhone 16 Pro

the search may not return the expected result.

This is why choosing the correct query type is important.

Sometimes you want to search across multiple fields.

Imagine a movie database.

Each movie has:

• Title
• Description
• Genre

A user searches:

Adventure

Should Elasticsearch only search the title?

No.

It should search:

• Movie Title
• Movie Description
• Genre

A Multi Match Query searches several fields at the same time.

This helps users find relevant results even when the keyword appears in different parts of a document.

Suppose a customer searches:

Gaming Laptop

Elasticsearch analyzes both words separately.

It searches for:

• Gaming
• Laptop

Then combines the results to find the best matching products.

This makes searches much more flexible than exact phrase matching.

Suppose a customer searches:

Purple Dragon Laptop

No such product exists.

Instead of displaying incorrect information, Elasticsearch simply returns:

No matching documents found.

Many applications then display messages such as:

• No products found.
• Try another keyword.
• Did you mean...?

These features improve the user experience.

Modern applications often show results while you're still typing.

For example:

User types:

Lap

Possible suggestions:

• Laptop
• Laptop Bag
• Laptop Stand

This feature is called Autocomplete.

Elasticsearch supports autocomplete using specialized indexing techniques, allowing users to find information more quickly.

Imagine searching on Netflix.

You type:

Avengers

Immediately, Netflix displays:

• Avengers
• Avengers: Endgame
• Avengers: Infinity War

This fast response is possible because Elasticsearch (or similar search technologies) uses optimized search indices instead of scanning every movie individually.

Many beginners confuse searching and filtering.

Although they seem similar, they serve different purposes.

Searching

Searching tries to find the most relevant information.

Example:

Search:

Gaming Laptop

Elasticsearch looks for documents that best match those words.

Filtering

Filtering narrows down the search results using specific conditions.

Example:

After searching for laptops, a customer selects:

• Brand: Dell
• Price: Under ₹60,000
• Rating: Above 4 Stars

These conditions filter the results.

Unlike searching, filtering usually does not calculate relevance scores.

Imagine an online shopping website.

A customer searches:

Laptop

10,000 products match.

The customer then applies filters:

• Dell
• 16 GB RAM
• SSD Storage
• Under ₹70,000

Now only 45 products remain.

Searching finds possible matches.

Filtering narrows those matches.

Traditional databases mainly compare values.

Elasticsearch goes much further by:

• Understanding analyzed text.
• Ranking results by relevance.
• Supporting full-text search.
• Searching across multiple fields.
• Handling millions of documents efficiently.
• Returning results in milliseconds.

This is why many modern applications use Elasticsearch for their search functionality.

• Use Match Queries for normal text searches.
• Use Term Queries when searching for exact values.
• Use Multi Match Queries when searching across multiple fields.
• Filtering is used to narrow results after or alongside a search.
• Elasticsearch ranks search results automatically based on relevance.