Редагувати

Поділитися через


Quickstart: Vector search by using REST

Learn how to use the Search REST APIs to create, load, and query vectors in Azure AI Search.

In Azure AI Search, a vector store has an index schema that defines vector and nonvector fields, a vector configuration for algorithms that create the embedding space, and settings on vector field definitions that are used in query requests. The Create Index API creates the vector store.

If you don't have an Azure subscription, create a free account before you begin.

Note

This quickstart omits the vectorization step and provides embeddings in sample documents. If you want to add built-in data chunking and vectorization over your own content, try the Import and vectorize data wizard for an end-to-end walkthrough.

Prerequisites

  • Visual Studio Code with a REST client. If you need help with getting started, see Quickstart: Text search using REST.

  • Azure AI Search, in any region and on any tier. You can use the Free tier for this quickstart, but Basic or higher is recommended for larger data files. Create or find an existing Azure AI Search resource under your current subscription.

    Most existing services support vector search. For a small subset of services created prior to January 2019, an index that contains vector fields fails on creation. In this situation, a new service must be created.

  • Optionally, to run the query example that invokes semantic reranking, your search service must be the Basic tier or higher, with semantic ranker enabled.

  • Optionally, an Azure OpenAI resource with a deployment of text-embedding-ada-002. The source .rest file includes an optional step for generating new text embeddings, but we provide pregenerated embeddings so that you can omit this dependency.

Download files

Download a REST sample from GitHub to send the requests in this quickstart. For more information, see Downloading files from GitHub.

You can also start a new file on your local system and create requests manually by using the instructions in this article.

Get a search service endpoint

You can find the search service endpoint in the Azure portal.

  1. Sign in to the Azure portal and find your search service.

  2. On the Overview home page, find the URL. An example endpoint might look like https://mydemo.search.windows.net.

    Screenshot of the URL property on the overview page.

You're pasting this endpoint into the .rest or .http file in a later step.

Configure access

Requests to the search endpoint must be authenticated and authorized. You can use API keys or roles for this task. Keys are easier to start with, but roles are more secure.

For a role-based connection, the following instructions have you connecting to Azure AI Search under your identity, not the identity of a client app.

Option 1: Use keys

Select Settings > Keys and then copy an admin key. Admin keys are used to add, modify, and delete objects. There are two interchangeable admin keys. Copy either one. For more information, see Connect to Azure AI Search using key authentication.

Screenshot that shows the API keys in the Azure portal.

You're pasting this key into the .rest or .http file in a later step.

Option 2: Use roles

Make sure your search service is configured for role-based access. You must have preconfigured role-assignments for developer access. Your role assignments must grant permission to create, load, and query a search index.

In this section, obtain your personal identity token using either the Azure CLI, Azure PowerShell, or the Azure portal.

  1. Sign in to Azure CLI.

    az login
    
  2. Get your personal identity token.

    az account get-access-token --scope https://search.azure.com/.default
    

You're pasting your personal identity token into the .rest or .http file in a later step.

Note

This section assumes you're using a local client that connects to Azure AI Search on your behalf. An alternative approach is getting a token for the client app, assuming your application is registered with Microsoft Entra ID.

Create a vector index

Create Index (REST) creates a vector index and sets up the physical data structures on your search service.

The index schema is organized around hotel content. Sample data consists of vector and nonvector names and descriptions of seven fictitious hotels. This schema includes configurations for vector indexing and queries, and for semantic ranking.

  1. Open a new text file in Visual Studio Code.

  2. Set variables to the values you collected earlier. This example uses a personal identity token.

    @baseUrl = PUT-YOUR-SEARCH-SERVICE-URL-HERE
    @token = PUT-YOUR-PERSONAL-IDENTITY-TOKEN-HERE
    
  3. Save the file with a .rest or .http file extension.

  4. Paste in the following example to create the hotels-vector-quickstart index on your search service.

    ### Create a new index
    POST {{baseUrl}}/indexes?api-version=2023-11-01  HTTP/1.1
        Content-Type: application/json
        Authorization: Bearer {{token}}
    
    {
        "name": "hotels-vector-quickstart",
        "fields": [
            {
                "name": "HotelId", 
                "type": "Edm.String",
                "searchable": false, 
                "filterable": true, 
                "retrievable": true, 
                "sortable": false, 
                "facetable": false,
                "key": true
            },
            {
                "name": "HotelName", 
                "type": "Edm.String",
                "searchable": true, 
                "filterable": false, 
                "retrievable": true, 
                "sortable": true, 
                "facetable": false
            },
            {
                "name": "HotelNameVector",
                "type": "Collection(Edm.Single)",
                "searchable": true,
                "retrievable": true,
                "dimensions": 1536,
                "vectorSearchProfile": "my-vector-profile"
            },
            {
                "name": "Description", 
                "type": "Edm.String",
                "searchable": true, 
                "filterable": false, 
                "retrievable": true, 
                "sortable": false, 
                "facetable": false
            },
            {
                "name": "DescriptionVector",
                "type": "Collection(Edm.Single)",
                "searchable": true,
                "retrievable": true,
                "dimensions": 1536,
                "vectorSearchProfile": "my-vector-profile"
            },
            {
                "name": "Category", 
                "type": "Edm.String",
                "searchable": true, 
                "filterable": true, 
                "retrievable": true, 
                "sortable": true, 
                "facetable": true
            },
            {
                "name": "Tags",
                "type": "Collection(Edm.String)",
                "searchable": true,
                "filterable": true,
                "retrievable": true,
                "sortable": false,
                "facetable": true
            },
            {
                "name": "Address", 
                "type": "Edm.ComplexType",
                "fields": [
                    {
                        "name": "City", "type": "Edm.String",
                        "searchable": true, "filterable": true, "retrievable": true, "sortable": true, "facetable": true
                    },
                    {
                        "name": "StateProvince", "type": "Edm.String",
                        "searchable": true, "filterable": true, "retrievable": true, "sortable": true, "facetable": true
                    }
                ]
            },
            {
                "name": "Location",
                "type": "Edm.GeographyPoint",
                "searchable": false, 
                "filterable": true, 
                "retrievable": true, 
                "sortable": true, 
                "facetable": false
            }
        ],
        "vectorSearch": {
            "algorithms": [
                {
                    "name": "my-hnsw-vector-config-1",
                    "kind": "hnsw",
                    "hnswParameters": 
                    {
                        "m": 4,
                        "efConstruction": 400,
                        "efSearch": 500,
                        "metric": "cosine"
                    }
                },
                {
                    "name": "my-hnsw-vector-config-2",
                    "kind": "hnsw",
                    "hnswParameters": 
                    {
                        "m": 4,
                        "metric": "euclidean"
                    }
                },
                {
                    "name": "my-eknn-vector-config",
                    "kind": "exhaustiveKnn",
                    "exhaustiveKnnParameters": 
                    {
                        "metric": "cosine"
                    }
                }
            ],
            "profiles": [      
                {
                    "name": "my-vector-profile",
                    "algorithm": "my-hnsw-vector-config-1"
                }
          ]
        },
        "semantic": {
            "configurations": [
                {
                    "name": "my-semantic-config",
                    "prioritizedFields": {
                        "titleField": {
                            "fieldName": "HotelName"
                        },
                        "prioritizedContentFields": [
                            { "fieldName": "Description" }
                        ],
                        "prioritizedKeywordsFields": [
                            { "fieldName": "Tags" }
                        ]
                    }
                }
            ]
        }
    }
    
  5. Select Send request. Recall that you need the REST client to send requests. You should have an HTTP/1.1 201 Created response. The response body should include the JSON representation of the index schema.

    Key points:

    • The fields collection includes a required key field and text and vector fields (such as Description and DescriptionVector) for text and vector search. Colocating vector and nonvector fields in the same index enables hybrid queries. For instance, you can combine filters, text search with semantic ranking, and vectors into a single query operation.
    • Vector fields must be type: Collection(Edm.Single) with dimensions and vectorSearchProfile properties.
    • The vectorSearch section is an array of approximate nearest neighbor algorithm configurations and profiles. Supported algorithms include hierarchical navigable small world and exhaustive k-nearest neighbor. For more information, see Relevance scoring in vector search.
    • [Optional]: The semantic configuration enables reranking of search results. You can rerank results in queries of type semantic for string fields that are specified in the configuration. To learn more, see Semantic ranking overview.

Upload documents

Creating and loading the index are separate steps. In Azure AI Search, the index contains all searchable data and queries run on the search service. For REST calls, the data is provided as JSON documents. Use Documents- Index REST API for this task.

The URI is extended to include the docs collection and the index operation.

Important

The following example isn't runnable code. For readability, we excluded vector values because each one contains 1,536 embeddings, which is too long for this article. If you want to try this step, copy runnable code from the sample on GitHub.

### Upload documents
POST {{baseUrl}}/indexes/hotels-quickstart-vectors/docs/index?api-version=2023-11-01  HTTP/1.1
Content-Type: application/json
Authorization: Bearer {{token}}

{
    "value": [
        {
            "@search.action": "mergeOrUpload",
            "HotelId": "1",
            "HotelName": "Stay-Kay City Hotel",
            "HotelNameVector": [VECTOR ARRAY OMITTED],
            "Description": 
                "The hotel is ideally located on the main commercial artery of the city 
                in the heart of New York.",
            "DescriptionVector": [VECTOR ARRAY OMITTED],
            "Category": "Boutique",
            "Tags": [
                "pool",
                "air conditioning",
                "concierge"
            ],
        },
        {
            "@search.action": "mergeOrUpload",
            "HotelId": "2",
            "HotelName": "Old Century Hotel",
            "HotelNameVector": [VECTOR ARRAY OMITTED],
            "Description": 
                "The hotel is situated in a  nineteenth century plaza, which has been 
                expanded and renovated to the highest architectural standards to create a modern, 
                functional and first-class hotel in which art and unique historical elements 
                coexist with the most modern comforts.",
            "DescriptionVector": [VECTOR ARRAY OMITTED],
            "Category": "Boutique",
            "Tags": [
                "pool",
                "air conditioning",
                "free wifi",
                "concierge"
            ]
        },
        {
            "@search.action": "mergeOrUpload",
            "HotelId": "3",
            "HotelName": "Gastronomic Landscape Hotel",
            "HotelNameVector": [VECTOR ARRAY OMITTED],
            "Description": 
                "The Hotel stands out for its gastronomic excellence under the management of 
                William Dough, who advises on and oversees all of the Hotel’s restaurant services.",
            "DescriptionVector": [VECTOR ARRAY OMITTED],
            "Category": "Resort and Spa",
            "Tags": [
                "air conditioning",
                "bar",
                "continental breakfast"
            ]
        }
        {
            "@search.action": "mergeOrUpload",
            "HotelId": "4",
            "HotelName": "Sublime Palace Hotel",
            "HotelNameVector": [VECTOR ARRAY OMITTED],
            "Description": 
                "Sublime Palace Hotel is located in the heart of the historic center of 
                Sublime in an extremely vibrant and lively area within short walking distance to 
                the sites and landmarks of the city and is surrounded by the extraordinary beauty 
                of churches, buildings, shops and monuments. 
                Sublime Palace is part of a lovingly restored 1800 palace.",
            "DescriptionVector": [VECTOR ARRAY OMITTED],
            "Category": "Boutique",
            "Tags": [
                "concierge",
                "view",
                "24-hour front desk service"
            ]
        },
        {
            "@search.action": "mergeOrUpload",
            "HotelId": "13",
            "HotelName": "Luxury Lion Resort",
            "HotelNameVector": [VECTOR ARRAY OMITTED],
            "Description": 
                "Unmatched Luxury.  Visit our downtown hotel to indulge in luxury 
                accommodations. Moments from the stadium, we feature the best in comfort",
            "DescriptionVector": [VECTOR ARRAY OMITTED],
            "Category": "Resort and Spa",
            "Tags": [
                "view",
                "free wifi",
                "pool"
            ]
        },
        {
            "@search.action": "mergeOrUpload",
            "HotelId": "48",
            "HotelName": "Nordick's Valley Motel",
            "HotelNameVector": [VECTOR ARRAY OMITTED],
            "Description": 
                "Only 90 miles (about 2 hours) from the nation's capital and nearby 
                most everything the historic valley has to offer.  Hiking? Wine Tasting? Exploring 
                the caverns?  It's all nearby and we have specially priced packages to help make 
                our B&B your home base for fun while visiting the valley.",
            "DescriptionVector": [VECTOR ARRAY OMITTED],
            "Category": "Boutique",
            "Tags": [
                "continental breakfast",
                "air conditioning",
                "free wifi"
            ],
        },
        {
            "@search.action": "mergeOrUpload",
            "HotelId": "49",
            "HotelName": "Swirling Currents Hotel",
            "HotelNameVector": [VECTOR ARRAY OMITTED],
            "Description": 
                "Spacious rooms, glamorous suites and residences, rooftop pool, walking 
                access to shopping, dining, entertainment and the city center.",
            "DescriptionVector": [VECTOR ARRAY OMITTED],
            "Category": "Luxury",
            "Tags": [
                "air conditioning",
                "laundry service",
                "24-hour front desk service"
            ]
        }
    ]
}

Key points:

  • Documents in the payload consist of fields defined in the index schema.
  • Vector fields contain floating point values. The dimensions attribute has a minimum of 2 and a maximum of 3,072 floating point values each. This quickstart sets the dimensions attribute to 1,536 because that's the size of embeddings generated by the Azure OpenAI text-embedding-ada-002 model.

Run queries

Now that documents are loaded, you can issue vector queries against them by using Documents - Search Post (REST).

There are several queries to demonstrate various patterns:

The vector queries in this section are based on two strings:

  • Search string: historic hotel walk to restaurants and shopping
  • Vector query string (vectorized into a mathematical representation): classic lodging near running trails, eateries, retail

The vector query string is semantically similar to the search string, but it includes terms that don't exist in the search index. If you do a keyword search for classic lodging near running trails, eateries, retail, results are zero. We use this example to show how you can get relevant results even if there are no matching terms.

Important

The following examples aren't runnable code. For readability, we excluded vector values because each array contains 1,536 embeddings, which is too long for this article. If you want to try these queries, copy runnable code from the sample on GitHub.

  1. Paste in a POST request to query the search index. Then select Send request. The URI is extended to include the /docs/search operator.

    ### Run a query
    POST {{baseUrl}}/indexes/hotels-vector-quickstart/docs/search?api-version=2023-11-01  HTTP/1.1
        Content-Type: application/json
        Authorization: Bearer {{token}}
    
        {
            "count": true,
            "select": "HotelId, HotelName, Description, Category",
            "vectorQueries": [
                {
                    "vector"": [0.01944167, 0.0040178085
                        . . .  TRIMMED FOR BREVITY
                        010858015, -0.017496133],
                    "k": 7,
                    "fields": "DescriptionVector",
                    "kind": "vector",
                    "exhaustive": true
                }
            ]
        }
    

    This vector query is shortened for brevity. The vectorQueries.vector contains the vectorized text of the query input, fields determines which vector fields are searched, and k specifies the number of nearest neighbors to return.

    The vector query string is classic lodging near running trails, eateries, retail, which is vectorized into 1,536 embeddings for this query.

  2. Review the response. The response for the vector equivalent of classic lodging near running trails, eateries, retail includes seven results. Each result provides a search score and the fields listed in select. In a similarity search, the response always includes k results ordered by the value similarity score.

    {
        "@odata.context": "https://my-demo-search.search.windows.net/indexes('hotels-vector-quickstart')/$metadata#docs(*)",
        "@odata.count": 7,
        "value": [
            {
                "@search.score": 0.857736,
                "HotelName": "Nordick's Valley Motel",
                "Description": "Only 90 miles (about 2 hours) from the nation's capital and nearby most everything the historic valley has to offer.  Hiking? Wine Tasting? Exploring the caverns?  It's all nearby and we have specially priced packages to help make our B&B your home base for fun while visiting the valley."
            },
            {
                "@search.score": 0.8399129,
                "HotelName": "Swirling Currents Hotel",
                "Description": "Spacious rooms, glamorous suites and residences, rooftop pool, walking access to shopping, dining, entertainment and the city center."
            },
            {
                "@search.score": 0.8383954,
                "HotelName": "Luxury Lion Resort",
                "Description": "Unmatched Luxury.  Visit our downtown hotel to indulge in luxury accommodations. Moments from the stadium, we feature the best in comfort"
            },
            {
                "@search.score": 0.8254346,
                "HotelName": "Sublime Palace Hotel",
                "Description": "Sublime Palace Hotel is located in the heart of the historic center of Sublime in an extremely vibrant and lively area within short walking distance to the sites and landmarks of the city and is surrounded by the extraordinary beauty of churches, buildings, shops and monuments. Sublime Palace is part of a lovingly restored 1800 palace."
            },
            {
                "@search.score": 0.82380056,
                "HotelName": "Stay-Kay City Hotel",
                "Description": "The hotel is ideally located on the main commercial artery of the city in the heart of New York."
            },
            {
                "@search.score": 0.81514084,
                "HotelName": "Old Century Hotel",
                "Description": "The hotel is situated in a  nineteenth century plaza, which has been expanded and renovated to the highest architectural standards to create a modern, functional and first-class hotel in which art and unique historical elements coexist with the most modern comforts."
            },
            {
                "@search.score": 0.8133763,
                "HotelName": "Gastronomic Landscape Hotel",
                "Description": "The Hotel stands out for its gastronomic excellence under the management of William Dough, who advises on and oversees all of the Hotel’s restaurant services."
            }
        ]
    }
    

Single vector search with filter

You can add filters, but the filters are applied to the nonvector content in your index. In this example, the filter applies to the Tags field to filter out any hotels that don't provide free Wi-Fi.

  1. Paste in a POST request to query the search index.

    ### Run a vector query with a filter
    POST {{baseUrl}}/indexes/hotels-vector-quickstart/docs/search?api-version=2023-11-01  HTTP/1.1
        Content-Type: application/json
        Authorization: Bearer {{token}}
    
        {
            "count": true,
            "select": "HotelId, HotelName, Category, Tags, Description",
            "filter": "Tags/any(tag: tag eq 'free wifi')",
            "vectorFilterMode": "postFilter",
            "vectorQueries": [
            {
                "vector": [ VECTOR OMITTED ],
                "k": 7,
                "fields": "DescriptionVector",
                "kind": "vector",
                "exhaustive": true
            },
        ]
    }
    
  2. Review the response. The query is the same as the previous example, but it includes a post-processing exclusion filter and returns only the three hotels that have free Wi-Fi.

    {
    
        "@odata.count": 3,
        "value": [
            {
                "@search.score": 0.857736,
                "HotelName": "Nordick's Valley Motel",
                "Description": "Only 90 miles (about 2 hours) from the nation's capital and nearby most everything the historic valley has to offer.  Hiking? Wine Tasting? Exploring the caverns?  It's all nearby and we have specially priced packages to help make our B&B your home base for fun while visiting the valley.",
                "Tags": [
                    "continental breakfast",
                    "air conditioning",
                    "free wifi"
                ]
            },
            {
                "@search.score": 0.8383954,
                "HotelName": "Luxury Lion Resort",
                "Description": "Unmatched Luxury.  Visit our downtown hotel to indulge in luxury accommodations. Moments from the stadium, we feature the best in comfort",
                "Tags": [
                    "view",
                    "free wifi",
                    "pool"
                ]
            },
            {
                "@search.score": 0.81514084,
                "HotelName": "Old Century Hotel",
                "Description": "The hotel is situated in a  nineteenth century plaza, which has been expanded and renovated to the highest architectural standards to create a modern, functional and first-class hotel in which art and unique historical elements coexist with the most modern comforts.",
                "Tags": [
                    "pool",
                    "free wifi",
                    "concierge"
                ]
            }
        ]
    }
    

Hybrid search consists of keyword queries and vector queries in a single search request. This example runs the vector query and full text search concurrently:

  • Search string: historic hotel walk to restaurants and shopping
  • Vector query string (vectorized into a mathematical representation): classic lodging near running trails, eateries, retail
  1. Paste in a POST request to query the search index. Then select Send request.

    ### Run a hybrid query
    POST {{baseUrl}}/indexes/hotels-vector-quickstart/docs/search?api-version=2023-11-01  HTTP/1.1
        Content-Type: application/json
        Authorization: Bearer {{token}}
    
    {
        "count": true,
        "search": "historic hotel walk to restaurants and shopping",
        "select": "HotelName, Description",
        "top": 7,
        "vectorQueries": [
            {
                "vector": [ VECTOR OMITTED],
                "k": 7,
                "fields": "DescriptionVector",
                "kind": "vector",
                "exhaustive": true
            }
        ]
    }
    

    Because this is a hybrid query, results are ranked by Reciprocal Rank Fusion (RRF). RRF evaluates search scores of multiple search results, takes the inverse, and then merges and sorts the combined results. The top number of results are returned.

  2. Review the response.

    {
        "@odata.count": 7,
        "value": [
            {
                "@search.score": 0.03279569745063782,
                "HotelName": "Luxury Lion Resort",
                "Description": "Unmatched Luxury.  Visit our downtown hotel to indulge in luxury accommodations. Moments from the stadium, we feature the best in comfort"
            },
            {
                "@search.score": 0.03226646035909653,
                "HotelName": "Sublime Palace Hotel",
                "Description": "Sublime Palace Hotel is located in the heart of the historic center of Sublime in an extremely vibrant and lively area within short walking distance to the sites and landmarks of the city and is surrounded by the extraordinary beauty of churches, buildings, shops and monuments. Sublime Palace is part of a lovingly restored 1800 palace."
            },
            {
                "@search.score": 0.03226646035909653,
                "HotelName": "Swirling Currents Hotel",
                "Description": "Spacious rooms, glamorous suites and residences, rooftop pool, walking access to shopping, dining, entertainment and the city center."
            },
            {
                "@search.score": 0.03205128386616707,
                "HotelName": "Nordick's Valley Motel",
                "Description": "Only 90 miles (about 2 hours) from the nation's capital and nearby most everything the historic valley has to offer.  Hiking? Wine Tasting? Exploring the caverns?  It's all nearby and we have specially priced packages to help make our B&B your home base for fun while visiting the valley."
            },
            {
                "@search.score": 0.03128054738044739,
                "HotelName": "Gastronomic Landscape Hotel",
                "Description": "The Hotel stands out for its gastronomic excellence under the management of William Dough, who advises on and oversees all of the Hotel’s restaurant services."
            },
            {
                "@search.score": 0.03100961446762085,
                "HotelName": "Old Century Hotel",
                "Description": "The hotel is situated in a  nineteenth century plaza, which has been expanded and renovated to the highest architectural standards to create a modern, functional and first-class hotel in which art and unique historical elements coexist with the most modern comforts."
            },
            {
                "@search.score": 0.03077651560306549,
                "HotelName": "Stay-Kay City Hotel",
                "Description": "The hotel is ideally located on the main commercial artery of the city in the heart of New York."
            }
        ]
    }
    

    Because RRF merges results, it helps to review the inputs. The following results are from only the full-text query. The top two results are Sublime Palace Hotel and History Lion Resort. The Sublime Palace Hotel has a stronger BM25 relevance score.

            {
                "@search.score": 2.2626662,
                "HotelName": "Sublime Palace Hotel",
                "Description": "Sublime Palace Hotel is located in the heart of the historic center of Sublime in an extremely vibrant and lively area within short walking distance to the sites and landmarks of the city and is surrounded by the extraordinary beauty of churches, buildings, shops and monuments. Sublime Palace is part of a lovingly restored 1800 palace."
            },
            {
                "@search.score": 0.86421645,
                "HotelName": "Luxury Lion Resort",
                "Description": "Unmatched Luxury.  Visit our downtown hotel to indulge in luxury accommodations. Moments from the stadium, we feature the best in comfort"
                },
    

    In the vector-only query, which uses HNSW for finding matches, the Sublime Palace Hotel drops to fourth position. Historic Lion, which was second in the full-text search and third in the vector search, doesn't experience the same range of fluctuation, so it appears as a top match in a homogenized result set.

        "value": [
            {
                "@search.score": 0.857736,
                "HotelId": "48",
                "HotelName": "Nordick's Valley Motel",
                "Description": "Only 90 miles (about 2 hours) from the nation's capital and nearby most everything the historic valley has to offer.  Hiking? Wine Tasting? Exploring the caverns?  It's all nearby and we have specially priced packages to help make our B&B your home base for fun while visiting the valley.",
                "Category": "Boutique"
            },
            {
                "@search.score": 0.8399129,
                "HotelId": "49",
                "HotelName": "Swirling Currents Hotel",
                "Description": "Spacious rooms, glamorous suites and residences, rooftop pool, walking access to shopping, dining, entertainment and the city center.",
                "Category": "Luxury"
            },
            {
                "@search.score": 0.8383954,
                "HotelId": "13",
                "HotelName": "Luxury Lion Resort",
                "Description": "Unmatched Luxury.  Visit our downtown hotel to indulge in luxury accommodations. Moments from the stadium, we feature the best in comfort",
                "Category": "Resort and Spa"
            },
            {
                "@search.score": 0.8254346,
                "HotelId": "4",
                "HotelName": "Sublime Palace Hotel",
                "Description": "Sublime Palace Hotel is located in the heart of the historic center of Sublime in an extremely vibrant and lively area within short walking distance to the sites and landmarks of the city and is surrounded by the extraordinary beauty of churches, buildings, shops and monuments. Sublime Palace is part of a lovingly restored 1800 palace.",
                "Category": "Boutique"
            },
            {
                "@search.score": 0.82380056,
                "HotelId": "1",
                "HotelName": "Stay-Kay City Hotel",
                "Description": "The hotel is ideally located on the main commercial artery of the city in the heart of New York.",
                "Category": "Boutique"
            },
            {
                "@search.score": 0.81514084,
                "HotelId": "2",
                "HotelName": "Old Century Hotel",
                "Description": "The hotel is situated in a  nineteenth century plaza, which has been expanded and renovated to the highest architectural standards to create a modern, functional and first-class hotel in which art and unique historical elements coexist with the most modern comforts.",
                "Category": "Boutique"
            },
            {
                "@search.score": 0.8133763,
                "HotelId": "3",
                "HotelName": "Gastronomic Landscape Hotel",
                "Description": "The Hotel stands out for its gastronomic excellence under the management of William Dough, who advises on and oversees all of the Hotel’s restaurant services.",
                "Category": "Resort and Spa"
            }
        ]
    

Semantic hybrid search with a filter

Here's the last query in the collection. This hybrid query with semantic ranking is filtered to show only the hotels within a 500-kilometer radius of Washington D.C. You can set vectorFilterMode to null, which is equivalent to the default (preFilter for newer indexes and postFilter for older ones).

  1. Paste in a POST request to query the search index. Then select Send request.

    ### Run a hybrid query
    POST {{baseUrl}}/indexes/hotels-vector-quickstart/docs/search?api-version=2023-11-01  HTTP/1.1
        Content-Type: application/json
        Authorization: Bearer {{token}}
    
    {
        "count": true,
        "search": "historic hotel walk to restaurants and shopping",
        "select": "HotelId, HotelName, Category, Description,Address/City, Address/StateProvince",
        "filter": "geo.distance(Location, geography'POINT(-77.03241 38.90166)') le 500",
        "vectorFilterMode": null,
        "facets": [ "Address/StateProvince"],
        "top": 7,
        "queryType": "semantic",
        "answers": "extractive|count-3",
        "captions": "extractive|highlight-true",
        "semanticConfiguration": "my-semantic-config",
        "vectorQueries": [
            {
                "vector": [ VECTOR OMITTED ],
                "k": 7,
                "fields": "DescriptionVector",
                "kind": "vector",
                "exhaustive": true
            }
        ]
    }
    
  2. Review the response. The response is three hotels, which are filtered by location and faceted by StateProvince and semantically reranked to promote results that are closest to the search string query (historic hotel walk to restaurants and shopping).

    The Swirling Currents Hotel now moves into the top spot. Without semantic ranking, Nordick's Valley Motel is number one. With semantic ranking, the machine comprehension models recognize that historic applies to "hotel, within walking distance to dining (restaurants) and shopping."

    {
        "@odata.count": 3,
        "@search.facets": {
            "Address/StateProvince": [
                {
                    "count": 1,
                    "value": "NY"
                },
                {
                    "count": 1,
                    "value": "VA"
                }
            ]
        },
        "@search.answers": [],
        "value": [
            {
                "@search.score": 0.03306011110544205,
                "@search.rerankerScore": 2.5094974040985107,
                "HotelId": "49",
                "HotelName": "Swirling Currents Hotel",
                "Description": "Spacious rooms, glamorous suites and residences, rooftop pool, walking access to shopping, dining, entertainment and the city center.",
                "Category": "Luxury",
                "Address": {
                    "City": "Arlington",
                    "StateProvince": "VA"
                }
            },
            {
                "@search.score": 0.03306011110544205,
                "@search.rerankerScore": 2.0370211601257324,
                "HotelId": "48",
                "HotelName": "Nordick's Valley Motel",
                "Description": "Only 90 miles (about 2 hours) from the nation's capital and nearby most everything the historic valley has to offer.  Hiking? Wine Tasting? Exploring the caverns?  It's all nearby and we have specially priced packages to help make our B&B your home base for fun while visiting the valley.",
                "Category": "Boutique",
                "Address": {
                    "City": "Washington D.C.",
                    "StateProvince": null
                }
            },
            {
                "@search.score": 0.032258063554763794,
                "@search.rerankerScore": 1.6706111431121826,
                "HotelId": "1",
                "HotelName": "Stay-Kay City Hotel",
                "Description": "The hotel is ideally located on the main commercial artery of the city in the heart of New York.",
                "Category": "Boutique",
                "Address": {
                    "City": "New York",
                    "StateProvince": "NY"
                }
            }
        ]
    }
    

    Key points:

    • Vector search is specified through the vectors.value property. Keyword search is specified through the search property.
    • In a hybrid search, you can integrate vector search with full-text search over keywords. Filters, spell check, and semantic ranking apply to textual content only, and not vectors. In this final query, there's no semantic answer because the system didn't produce one that was sufficiently strong.
    • Actual results include more detail, including semantic captions and highlights. Results were modified for readability. To get the full structure of the response, run the request in the REST client.

Clean up

When you're working in your own subscription, it's a good idea at the end of a project to identify whether you still need the resources you created. Resources left running can cost you money. You can delete resources individually or delete the resource group to delete the entire set of resources.

You can find and manage resources in the portal by using the All resources or Resource groups link in the leftmost pane.

You can also try this DELETE command:

### Delete an index
DELETE  {{baseUrl}}/indexes/hotels-vector-quickstart?api-version=2023-11-01 HTTP/1.1
    Content-Type: application/json
    Authorization: Bearer {{token}}

Next steps

As a next step, we recommend that you review the demo code for Python, C#, or JavaScript.