multi_match query clarification

Author: alexeygrigorev

01-intro/README.md

@@ -151,6 +151,8 @@ Query:
 }
 ```
 
+We use `"type": "best_fields"`. You can read more about 
+different types of `multi_match` search in [elastic-search.md](elastic-search.md).
 
 
 # Notes

01-intro/elastic-search.md

@@ -0,0 +1,107 @@
+# Elastic Search 
+
+This document contains useful things about Elasticsearch
+
+# `multi_match` Query in Elasticsearch
+
+The `multi_match` query is used to search for a given text across multiple fields in an Elasticsearch index.
+
+It provides various types to control how the matching is executed and scored. 
+
+There are multiple types of `multi_match` queries:
+
+- `best_fields`: Returns the highest score from any one field.
+- `most_fields`: Combines the scores from all fields.
+- `cross_fields`: Treats fields as one big field for scoring.
+- `phrase`: Searches for the query as an exact phrase.
+- `phrase_prefix`: Searches for the query as a prefix of a phrase.
+
+
+## `best_fields`
+
+The `best_fields` type searches each field separately and returns the highest score from any one of the fields.
+
+This type is useful when you want to find documents where at least one field matches the query well.
+
+
+```json
+{
+    "size": 5,
+    "query": {
+        "bool": {
+            "must": {
+                "multi_match": {
+                    "query": "How do I run docker on Windows?",
+                    "fields": ["question", "text"],
+                    "type": "best_fields"
+                }
+            }
+        }
+    }
+}
+```
+
+## `most_fields`
+
+The `most_fields` type searches each field and combines the scores from all fields.
+
+This is useful when the relevance of a document increases with more matching fields.
+
+```json
+{
+    "multi_match": {
+        "query": "How do I run docker on Windows?",
+        "fields": ["question^4", "text"],
+        "type": "most_fields"
+    }
+}
+```
+
+## `cross_fields`
+
+The `cross_fields` type treats fields as though they were one big field.
+
+It is suitable for cases where you have fields representing the same text in different ways, such as synonyms.
+
+```json
+{
+    "multi_match": {
+        "query": "How do I run docker on Windows?",
+        "fields": ["question", "text"],
+        "type": "cross_fields"
+    }
+}
+```
+
+## `phrase`
+
+The `phrase` type looks for the query as an exact phrase within the fields.
+
+It is useful for exact match searches.
+
+```json
+{
+    "multi_match": {
+        "query": "How do I run docker on Windows?",
+        "fields": ["question", "text"],
+        "type": "phrase"
+    }
+}
+```
+
+## `phrase_prefix`
+
+The `phrase_prefix` type searches for documents that contain the query as a prefix of a phrase.
+
+This is useful for autocomplete or typeahead functionality.
+
+
+```json
+{
+    "multi_match": {
+        "query": "How do I run docker on Windows?",
+        "fields": ["question", "text"],
+        "type": "phrase_prefix"
+    }
+}
+```

cohorts/2024/01-intro/homework.md

@@ -61,10 +61,12 @@ Which function do you use for adding your data to elastic?
 
 Now let's search in our index. 
 
-For a query "How do I execute a command in a running docker container?", what's the score for the top ranking result? 
+We will execute a query "How do I execute a command in a running docker container?". 
 
 Use only `question` and `text` fields and give `question` a boost of 4, and use `"type": "best_fields"`.
 
+What's the score for the top ranking result?
+
 * 94.05
 * 84.05
 * 74.05