Azure · ShivangiReja · Nov 7, 2023 · Nov 7, 2023 · Nov 7, 2023
@@ -1,6 +1,6 @@
 # Release History
 
-## 11.5.0-beta.6 (Unreleased)
+## 11.5.0 (Unreleased)
 
 ### Features Added
 

@@ -208,7 +208,9 @@ Semantic search enhances the quality of search results for text-based queries. B
 - It applies secondary ranking to the initial result set, promoting the most semantically relevant results to the top.
 - It extracts and returns captions and answers in the response, which can be displayed on a search page to enhance the user's search experience.
 
-To learn more about Semantic Search, you can refer to the [documentation](https://learn.microsoft.com/azure/search/vector-search-overview).
+To learn more about Semantic Search, you can refer to the [sample](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample08_SemanticSearch.md).
+
+Additionally, for more comprehensive information about Semantic Search, including its concepts and usage, you can refer to the [documentation](https://learn.microsoft.com/azure/search/semantic-search-overview). The documentation provides in-depth explanations and guidance on leveraging the power of Semantic Search in Azure Cognitive Search.
 
 #### Vector Search
 

@@ -2,5 +2,5 @@
   "AssetsRepo": "Azure/azure-sdk-assets",
   "AssetsRepoPrefixPath": "net",
   "TagPrefix": "net/search/Azure.Search.Documents",
-  "Tag": "net/search/Azure.Search.Documents_547641209a"
+  "Tag": "net/search/Azure.Search.Documents_c3ad072221"
 }
@@ -18,3 +18,4 @@ description: Samples for the Azure.Search.Documents client library
 - Learn about different ways to [index documents](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample05_IndexingDocuments.md).
 - Use your own Key Vault key to [encrypt your data source connections, skillsets, indexers, and index](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample06_EncryptedIndex.md).
 - Learn how to [index vector fields and perform vector search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch.md).
+- Learn how to [perform semantic search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample08_SemanticSearch.md).
@@ -102,7 +102,6 @@ SearchIndex index = new SearchIndex("movies")
 SearchableField genreField = new SearchableField("genre")
 {
     AnalyzerName = LexicalAnalyzerName.Values.EnLucene,
-    NormalizerName = LexicalNormalizerName.Lowercase,
     IsFacetable = true,
     IsFilterable = true
 };

@@ -8,17 +8,11 @@ Please refer the [documentation](https://learn.microsoft.com/azure/search/vector
 
 Here's the list of samples that will show you how to index the vector fields and perform vector search using .NET SDK.
 
-* [Vector Search Using RAW Vector Query](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingRawVectorQuery.md#vector-search-using-raw-vector-query)
-     * [Single Vector Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingRawVectorQuery.md#single-vector-search)
-     * [Single Vector Search With Filter](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingRawVectorQuery.md#single-vector-search-with-filter)
-     * [Hybrid Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingRawVectorQuery.md#hybrid-search)
-     * [Multi-vector Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingRawVectorQuery.md#multi-vector-search)
-     * [Multi-field Vector Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingRawVectorQuery.md#multi-field-vector-search)
-* [Vector Search Using Vectorizable Text Query](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizableTextQuery.md#vector-search-using-vectorizable-text-query)
-     * [Single Vector Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizableTextQuery.md#single-vector-search)
-     * [Single Vector Search With Filter](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizableTextQuery.md#single-vector-search-with-filter)
-     * [Hybrid Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizableTextQuery.md#hybrid-search)
-     * [Multi-vector Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizableTextQuery.md#multi-vector-search)
-     * [Multi-field Vector Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizableTextQuery.md#multi-field-vector-search)
+* [Vector Search Using Vectorized Query](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizedQuery.md#vector-search-using-vector-query)
+     * [Single Vector Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizedQuery.md#single-vector-search)
+     * [Single Vector Search With Filter](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizedQuery.md#single-vector-search-with-filter)
+     * [Hybrid Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizedQuery.md#hybrid-search)
+     * [Multi-Vector Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizedQuery.md#multi-vector-search)
+     * [Multi-Field Vector Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizedQuery.md#multi-field-vector-search)
 * [Vector Semantic Hybrid Search](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingSemanticHybridQuery.md)
 * [Vector Search Using Field Builder](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingFieldBuilder.md)
@@ -4,7 +4,7 @@ The `FieldBuilder` class allows you to define a Search index from a model type.
 
 ## Model Creation
 
-Consider the following model, which includes a property named `DescriptionVector` that represents a vector field. To configure a vector field, you must provide the model dimensions, indicating the size of the embeddings generated for this field, as well as the name of the vector search profile that specifies the algorithm configuration and any optional parameters for searching the vector field.
+Consider the following model, which includes a property named `DescriptionVector` that represents a vector field. To configure a vector field, you must provide the model dimensions, indicating the size of the embeddings generated for this field, as well as the name of the vector search profile that specifies the algorithm configuration in the `VectorSearchField` attribute.
 
 ```C# Snippet:Azure_Search_Documents_Tests_Samples_Sample07_Vector_Search_FieldBuilder_Model
 public class MyDocument
@@ -18,7 +18,7 @@ public class MyDocument
     [SearchableField(AnalyzerName = "en.microsoft")]
     public string Description { get; set; }
 
-    [SearchableField(VectorSearchDimensions = "1536", VectorSearchProfile = "my-vector-profile")]
+    [VectorSearchField(VectorSearchDimensions = 1536, VectorSearchProfileName = "my-vector-profile")]
     public IReadOnlyList<float> DescriptionVector { get; set; }
 }
 ```
@@ -28,7 +28,7 @@ public class MyDocument
 We will create an instace of `SearchIndex` and use `FieldBuilder` to define fields based on the `MyDocument` model class.
 
 ```C# Snippet:Azure_Search_Documents_Tests_Samples_Sample07_Vector_Search_Index_UsingFieldBuilder
-string vectorSearchProfile = "my-vector-profile";
+string vectorSearchProfileName = "my-vector-profile";
 string vectorSearchHnswConfig = "my-hsnw-vector-config";
 
 string indexName = "MyDocument";
@@ -40,11 +40,11 @@ SearchIndex searchIndex = new SearchIndex(indexName)
     {
         Profiles =
     {
-        new VectorSearchProfile(vectorSearchProfile, vectorSearchHnswConfig)
+        new VectorSearchProfile(vectorSearchProfileName, vectorSearchHnswConfig)
     },
         Algorithms =
     {
-        new HnswVectorSearchAlgorithmConfiguration(vectorSearchHnswConfig)
+        new HnswAlgorithmConfiguration(vectorSearchHnswConfig)
     }
     },
 };
@@ -61,4 +61,4 @@ SearchIndexClient indexClient = new(endpoint, credential);
 await indexClient.CreateIndexAsync(searchIndex);
 ```
 
-To perform vector search please refer to the [Vector Search Using RAW Vector Query](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingRawVectorQuery.md) or [Vector Search Using Vectorizable Text Query](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizableTextQuery.md) samples.
+To perform vector search please refer to the [Vector Search Using Vector Query](https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/search/Azure.Search.Documents/samples/Sample07_VectorSearch_UsingVectorizedQuery.md) sample.
@@ -9,7 +9,7 @@ Let's consider the example of a `Hotel`. First, we need to create an index for s
 We will create an instace of `SearchIndex` and define `Hotel` fields.
 
 ```C# Snippet:Azure_Search_Documents_Tests_Samples_Sample07_Vector_Semantic_Hybrid_Search_Index
-string vectorSearchProfile = "my-vector-profile";
+string vectorSearchProfileName = "my-vector-profile";
 string vectorSearchHnswConfig = "my-hsnw-vector-config";
 int modelDimensions = 1536;
 
@@ -21,47 +21,37 @@ SearchIndex searchIndex = new(indexName)
         new SimpleField("HotelId", SearchFieldDataType.String) { IsKey = true, IsFilterable = true, IsSortable = true, IsFacetable = true },
         new SearchableField("HotelName") { IsFilterable = true, IsSortable = true },
         new SearchableField("Description") { IsFilterable = true },
-        new SearchField("DescriptionVector", SearchFieldDataType.Collection(SearchFieldDataType.Single))
-        {
-            IsSearchable = true,
-            VectorSearchDimensions = modelDimensions,
-            VectorSearchProfile = vectorSearchProfile
-        },
+        new VectorSearchField("DescriptionVector", modelDimensions, vectorSearchProfileName),
         new SearchableField("Category") { IsFilterable = true, IsSortable = true, IsFacetable = true },
-        new SearchField("CategoryVector", SearchFieldDataType.Collection(SearchFieldDataType.Single))
-        {
-            IsSearchable = true,
-            VectorSearchDimensions = modelDimensions,
-            VectorSearchProfile = vectorSearchProfile
-        },
+        new VectorSearchField("CategoryVector", modelDimensions, vectorSearchProfileName),
     },
     VectorSearch = new()
     {
         Profiles =
         {
-            new VectorSearchProfile(vectorSearchProfile, vectorSearchHnswConfig)
+            new VectorSearchProfile(vectorSearchProfileName, vectorSearchHnswConfig)
         },
         Algorithms =
         {
-            new HnswVectorSearchAlgorithmConfiguration(vectorSearchHnswConfig)
+            new HnswAlgorithmConfiguration(vectorSearchHnswConfig)
         }
     },
-    SemanticSettings = new()
+    SemanticSearch = new()
     {
         Configurations =
-    {
-           new SemanticConfiguration("my-semantic-config", new()
-           {
-               TitleField = new(){ FieldName = "HotelName" },
-               ContentFields =
-               {
-                   new() { FieldName = "Description" }
-               },
-               KeywordFields =
-               {
-                   new() { FieldName = "Category" }
-               }
-           })
+        {
+            new SemanticConfiguration("my-semantic-config", new()
+            {
+                TitleField = new SemanticField("HotelName"),
+                ContentFields =
+                {
+                    new SemanticField("Description")
+                },
+                KeywordsFields =
+                {
+                    new SemanticField("Category")
+                }
+            })
         }
     }
 };
@@ -88,9 +78,9 @@ public class Hotel
     public string HotelId { get; set; }
     public string HotelName { get; set; }
     public string Description { get; set; }
-    public IReadOnlyList<float> DescriptionVector { get; set; }
+    public ReadOnlyMemory<float> DescriptionVector { get; set; }
     public string Category { get; set; }
-    public IReadOnlyList<float> CategoryVector { get; set; }
+    public ReadOnlyMemory<float> CategoryVector { get; set; }
 }
 ```
 
@@ -154,34 +144,43 @@ await searchClient.IndexDocumentsAsync(IndexDocumentsBatch.Upload(hotelDocuments
 
 ## Query Vector Data
 
-When using `RawVectorQuery`, the query for a vector field must also be a vector. To convert a text query string provided by a user into a vector representation, your application must call an embedding library that provides this capability. Use the same embedding library that you used to generate embeddings in the source documents. For more details on how to generate embeddings, please refer to the [documentation](https://learn.microsoft.com/azure/search/vector-search-how-to-generate-embeddings). In the sample codes below, we are using hardcoded embeddings to query vector field.
+When using `VectorizedQuery`, the query for a vector field must also be a vector. To convert a text query string provided by a user into a vector representation, your application must call an embedding library that provides this capability. Use the same embedding library that you used to generate embeddings in the source documents. For more details on how to generate embeddings, please refer to the [documentation](https://learn.microsoft.com/azure/search/vector-search-how-to-generate-embeddings). In the sample codes below, we are using hardcoded embeddings to query vector field.
 
 Let's query the index and make sure everything works as implemented. You can also refer to the [documentation](https://learn.microsoft.com/azure/search/vector-search-how-to-query?tabs=portal-vector-query#query-syntax-for-hybrid-search) for more information on querying vector data.
 
 ### Vector Semantic Hybrid Query
 
-In a vector semantic hybrid query, the `VectorQueries` collection contains the vectors representing the query input. The `Fields` property specifies which vector fields to search within. The `KNearestNeighborsCount` property dictates the number of nearest neighbors to return as top hits. With the semantic configuration added, we can proceed to execute a semantic hybrid query.
+In the context of vector search, the `Queries` collection contains the vectors that represent the query input. The `Fields` property specifies which vector fields should be searched. The `KNearestNeighborsCount` property determines the number of nearest neighbors to retrieve as the top hits.
+
+For semantic search, we will specify `SemanticSearch.SemanticConfigurationName` as `SearchQueryType.Semantic` in the `SearchOptions`. We will use the same `SemanticConfigurationName` that we defined when creating the index. Additionally, we have enabled `SemanticSearch.QueryCaption` and `SemanticSearch.QueryAnswer` in the `SearchOptions` to obtain the caption and answer in the response. With these configurations in place, we are prepared to execute a vector semantic hybrid query.
+
+With these settings in place, we're ready to execute a vector semantic hybrid query:
 
 ```C# Snippet:Azure_Search_Documents_Tests_Samples_Sample07_Vector_Semantic_Hybrid_Search
-IReadOnlyList<float> vectorizedResult = VectorSearchEmbeddings.SearchVectorizeDescription; // "Top hotels in town"
+ReadOnlyMemory<float> vectorizedResult = VectorSearchEmbeddings.SearchVectorizeDescription; // "Top hotels in town"
 
 SearchResults<Hotel> response = await searchClient.SearchAsync<Hotel>(
     "Is there any hotel located on the main commercial artery of the city in the heart of New York?",
-    new SearchOptions
-    {
-        VectorQueries = { new RawVectorQuery() { Vector = vectorizedResult, KNearestNeighborsCount = 3, Fields = { "DescriptionVector" } } },
-        QueryType = SearchQueryType.Semantic,
-        QueryLanguage = QueryLanguage.EnUs,
-        SemanticConfigurationName = "my-semantic-config",
-        QueryCaption = QueryCaptionType.Extractive,
-        QueryAnswer = QueryAnswerType.Extractive,
-    });
+     new SearchOptions
+     {
+         VectorSearch = new()
+         {
+             Queries = { new VectorizedQuery(vectorizedResult) { KNearestNeighborsCount = 3, Fields = { "DescriptionVector" } } }
+         },
+         SemanticSearch = new()
+         {
+             SemanticConfigurationName = "my-semantic-config",
+             QueryCaption = new(QueryCaptionType.Extractive),
+             QueryAnswer = new(QueryAnswerType.Extractive)
+         },
+         QueryType = SearchQueryType.Semantic,
+     });
 
 int count = 0;
 Console.WriteLine($"Semantic Hybrid Search Results:");
 
 Console.WriteLine($"\nQuery Answer:");
-foreach (AnswerResult result in response.Answers)
+foreach (QueryAnswerResult result in response.SemanticSearch.Answers)
 {
     Console.WriteLine($"Answer Highlights: {result.Highlights}");
     Console.WriteLine($"Answer Text: {result.Text}");
@@ -193,9 +192,9 @@ await foreach (SearchResult<Hotel> result in response.GetResultsAsync())
     Hotel doc = result.Document;
     Console.WriteLine($"{doc.HotelId}: {doc.HotelName}");
 
-    if (result.Captions != null)
+    if (result.SemanticSearch.Captions != null)
     {
-        var caption = result.Captions.FirstOrDefault();
+        var caption = result.SemanticSearch.Captions.FirstOrDefault();
         if (caption.Highlights != null && caption.Highlights != "")
         {
             Console.WriteLine($"Caption Highlights: {caption.Highlights}");