match, term and match_phrase

Key points

• match analyses your query text with the field analyzer, then looks up each resulting term.
• term performs no analysis at all, so it only ever finds a term exactly as it was indexed.
• That is why a term query on a text field with any capital letter usually returns zero hits.
• match_phrase requires the terms adjacent and in order, using positions from the postings.
• slop permits that many token moves, so a slop of 1 allows one word between the two terms.
• multi_match spreads one match across fields, with per-field boosts such as title^3.

Example

# title is mapped as text, so "Spring Boot" indexed as [spring, boot]

GET /articles/_search
{ "query": { "match": { "title": "Spring Boot" } } }        # 1 hit

GET /articles/_search
{ "query": { "term": { "title": "Spring Boot" } } }         # 0 hits

GET /articles/_search
{ "query": { "term": { "title.keyword": "Spring Boot" } } } # 1 hit

Use match on text fields and term on keyword fields - swapping them fails silently.