Search Basics — `_search`, `match`, and the Query DSL

What you'll walk away with

Explain the core ideas behind Search Basics — `_search`, `match`, and the Query DSL
Run the sample Elasticsearch query or code and verify its output
Apply the technique correctly to the Tutorial Platform and production scenarios

Build the mental model

All of Elasticsearch's search functionality goes through a single `_search` endpoint, where you write a JSON-based query language called the Query DSL (Domain Specific Language) in the request body. `match` is the most commonly used query type for full-text search — it tokenizes the search term through an analyzer (lowercasing, splitting words) and compares those tokens against the inverted index, so a search for "Redis Cache" can match a document containing just "redis" or just "cache" (OR logic by default). This differs sharply from a `term` query, which does exact, un-analyzed matching — `term` skips the analyzer and compares the raw value, so using it on a text field can silently miss many relevant results. The response's `hits.hits` array lists matching documents, each carrying a `_score` relevance score — a higher score means a closer match, and results are sorted by score, highest first, by default. Think of it like a Google search box: you type in keywords, and the engine ranks each document by how closely it matches before showing results.

Connect it to a real scenario

When a user types "redis cache" into the Tutorial Platform's search box, the frontend sends `GET /tutorials/_search` with a `match` query against the `title` field — using `term` instead would require an exact string match (spaces and capitalization included), badly hurting the experience. Render each search result card from `hits.hits[].source`'s tutorial title and summary, keeping them sorted by `_score` so the most relevant results appear first, matching what the user actually intended.

All search functionality goes through `_search` and a Query DSL JSON body
`match` is analyzed full-text search; `term` is exact, un-analyzed matching
The response's `_score` is the relevance ranking and the default sort key

Try the working example

http

GET /tutorials/_search
{
  "query": {
    "match": {
      "title": "redis cache"
    }
  }
}

You should see

You get a `hits.hits` array of tutorial documents whose `title` field contains "redis" or "cache", sorted by `_score`.

5-minute try-it

Write a request that searches for "database" on the `tags` field using `match`, then give an example scenario where searching the same term with `term` would produce a different result.

One important caution

Using a `term` query on a `text` field expecting an exact match — because it skips the analyzer, the indexed tokens (lowercased, stemmed) may not match the query value's original casing, returning zero results.

Assuming a multi-word `match` query requires every word to be present (AND logic) when its default behavior is actually OR logic.

Elasticsearch Guide — Search Your Data — Elastic

Easy traps

Using a `term` query on a `text` field expecting an exact match — because it skips the analyzer, the indexed tokens (lowercased, stemmed) may not match the query value's original casing, returning zero results.
Assuming a multi-word `match` query requires every word to be present (AND logic) when its default behavior is actually OR logic.
Validate sample queries and requests on a local or test instance with recoverable data before applying them to production.

Exercise

Write a request that searches for "database" on the `tags` field using `match`, then give an example scenario where searching the same term with `term` would produce a different result.

You'll know it worked when: You get a `hits.hits` array of tutorial documents whose `title` field contains "redis" or "cache", sorted by `_score`.