Azure Cognitive Search Full Text Search

Azure Cognitive Search:

Full-Text Search






Full-Text Search
Key components
Functional description
Query parsers
Separate query terms from query operators and create the query structure (a query tree) to be sent to the search engine.
Analyzers
Perform lexical analysis on query terms. This process can involve transforming, removing, or expanding of query terms.
Index
An efficient data structure used to store and organize searchable terms extracted from indexed documents.
Search engine
Retrieves and scores matching documents based on the contents of the inverted index.
Query Parsing
"search": "Spacious, air-condition* +\"Ocean view\"",


Term Query
Spacious


Phrase Query
"Ocean View"


Prefix Query
like air-condition


Lexical Analysis


Reducing a query term to the root form of a word


Removing non-essential words (​stopwords​, such as "the" or "and" in English)


Breaking a composite word into component parts


Lower casing an upper case word


Document Retrieval
{
"value": [
{
"id": "1",
"title": "Hotel Atman",
},
{
"id": "2",
"title": "Beach Resort",
},
{
"id": "3",
"title": "Playa Hotel",
},
{
"id": "4",
"title": "Ocean Retreat",
}
]
}
Inverted Index
Term
Document list
atman
1
beach
2
hotel
1, 3
Ocean
4
playa
3
resort
3
retreat
4
Scoring
{
"value": [
{
"@search.score": 0.25610128,
"id": "1",
"title": "Hotel Atman",
},
{
"@search.score": 0.08951007,
"id": "3",
"title": "Playa Hotel",
},
{
"@search.score": 0.05967338,
"id": "2",
"title": "Ocean Resort",
}
]
}


Relevance Tuning
Scoring Profiles
Promote documents in the ranked list of results based on a set of rules
Term Boosting
Provides a boosting operator ^ that can be applied to any part of the query tree.