Package dev.langchain4j.store.embedding.azure.search

Class AbstractAzureAiSearchEmbeddingStore

java.lang.Object

dev.langchain4j.store.embedding.azure.search.AbstractAzureAiSearchEmbeddingStore

All Implemented Interfaces:

dev.langchain4j.store.embedding.EmbeddingStore<dev.langchain4j.data.segment.TextSegment>

Direct Known Subclasses:

AzureAiSearchContentRetriever, AzureAiSearchEmbeddingStore

public abstract class AbstractAzureAiSearchEmbeddingStore
extends Object
implements dev.langchain4j.store.embedding.EmbeddingStore<dev.langchain4j.data.segment.TextSegment>

Field Summary

Fields

Modifier and Type	Field	Description
protected static final String	DEFAULT_FIELD_CONTENT
protected final String	DEFAULT_FIELD_CONTENT_VECTOR
protected static final String	DEFAULT_FIELD_METADATA
protected static final String	DEFAULT_FIELD_METADATA_ATTRS
protected static final String	DEFAULT_FIELD_METADATA_SOURCE
static final String	DEFAULT_INDEX_NAME
protected com.azure.search.documents.SearchClient	searchClient
protected static final String	SEMANTIC_SEARCH_CONFIG_NAME
protected static final String	VECTOR_ALGORITHM_NAME
protected static final String	VECTOR_SEARCH_PROFILE_NAME

Constructor Summary

Constructors

Constructor	Description
AbstractAzureAiSearchEmbeddingStore()

Method Summary

Modifier and Type	Method	Description
String	add(dev.langchain4j.data.embedding.Embedding embedding)	Add an embedding to the store.
String	add(dev.langchain4j.data.embedding.Embedding embedding, dev.langchain4j.data.segment.TextSegment textSegment)	Add an embedding and the related content to the store.
void	add(String id, dev.langchain4j.data.embedding.Embedding embedding)	Add an embedding to the store.
List<String>	addAll(List<dev.langchain4j.data.embedding.Embedding> embeddings)	Add a list of embeddings to the store.
List<String>	addAll(List<dev.langchain4j.data.embedding.Embedding> embeddings, List<dev.langchain4j.data.segment.TextSegment> embedded)	Add a list of embeddings, and the list of related content, to the store.
void	createOrUpdateIndex(int dimensions)	Creates or updates the index using a ready-made index.
void	deleteIndex()
List<dev.langchain4j.store.embedding.EmbeddingMatch<dev.langchain4j.data.segment.TextSegment>>	findRelevant(dev.langchain4j.data.embedding.Embedding referenceEmbedding, int maxResults, double minScore)
protected static double	fromAzureScoreToRelevanceScore(double score)	Calculates LangChain4j's RelevanceScore from Azure AI Search's score.
protected void	initialize(String endpoint, com.azure.core.credential.AzureKeyCredential keyCredential, com.azure.core.credential.TokenCredential tokenCredential, boolean createOrUpdateIndex, int dimensions, com.azure.search.documents.indexes.models.SearchIndex index, String indexName)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface dev.langchain4j.store.embedding.EmbeddingStore

findRelevant, findRelevant, findRelevant, remove, removeAll, removeAll, removeAll, search

Field Details

DEFAULT_INDEX_NAME

public static final String DEFAULT_INDEX_NAME

See Also:

• Constant Field Values

DEFAULT_FIELD_CONTENT

protected static final String DEFAULT_FIELD_CONTENT

See Also:

• Constant Field Values

DEFAULT_FIELD_CONTENT_VECTOR

protected final String DEFAULT_FIELD_CONTENT_VECTOR

See Also:

• Constant Field Values

DEFAULT_FIELD_METADATA

protected static final String DEFAULT_FIELD_METADATA

See Also:

• Constant Field Values

DEFAULT_FIELD_METADATA_SOURCE

protected static final String DEFAULT_FIELD_METADATA_SOURCE

See Also:

• Constant Field Values

DEFAULT_FIELD_METADATA_ATTRS

protected static final String DEFAULT_FIELD_METADATA_ATTRS

See Also:

• Constant Field Values

SEMANTIC_SEARCH_CONFIG_NAME

protected static final String SEMANTIC_SEARCH_CONFIG_NAME

See Also:

• Constant Field Values

VECTOR_ALGORITHM_NAME

protected static final String VECTOR_ALGORITHM_NAME

See Also:

• Constant Field Values

VECTOR_SEARCH_PROFILE_NAME

protected static final String VECTOR_SEARCH_PROFILE_NAME

See Also:

• Constant Field Values

searchClient

protected com.azure.search.documents.SearchClient searchClient

Constructor Details

AbstractAzureAiSearchEmbeddingStore

public AbstractAzureAiSearchEmbeddingStore()

Method Details

initialize

protected void initialize(String endpoint,
 com.azure.core.credential.AzureKeyCredential keyCredential,
 com.azure.core.credential.TokenCredential tokenCredential,
 boolean createOrUpdateIndex,
 int dimensions,
 com.azure.search.documents.indexes.models.SearchIndex index,
 String indexName)

createOrUpdateIndex

public void createOrUpdateIndex(int dimensions)

Creates or updates the index using a ready-made index.

Parameters:

dimensions - The number of dimensions of the embeddings.

deleteIndex

public void deleteIndex()

add

public String add(dev.langchain4j.data.embedding.Embedding embedding)

Add an embedding to the store.

Specified by:

add in interface dev.langchain4j.store.embedding.EmbeddingStore<dev.langchain4j.data.segment.TextSegment>

add

public void add(String id,
 dev.langchain4j.data.embedding.Embedding embedding)

Add an embedding to the store.

Specified by:

add in interface dev.langchain4j.store.embedding.EmbeddingStore<dev.langchain4j.data.segment.TextSegment>

add

public String add(dev.langchain4j.data.embedding.Embedding embedding,
 dev.langchain4j.data.segment.TextSegment textSegment)

Add an embedding and the related content to the store.

Specified by:

add in interface dev.langchain4j.store.embedding.EmbeddingStore<dev.langchain4j.data.segment.TextSegment>

addAll

public List<String> addAll(List<dev.langchain4j.data.embedding.Embedding> embeddings)

Add a list of embeddings to the store.

Specified by:

addAll in interface dev.langchain4j.store.embedding.EmbeddingStore<dev.langchain4j.data.segment.TextSegment>

addAll

public List<String> addAll(List<dev.langchain4j.data.embedding.Embedding> embeddings,
 List<dev.langchain4j.data.segment.TextSegment> embedded)

Add a list of embeddings, and the list of related content, to the store.

Specified by:

addAll in interface dev.langchain4j.store.embedding.EmbeddingStore<dev.langchain4j.data.segment.TextSegment>

findRelevant

public List<dev.langchain4j.store.embedding.EmbeddingMatch<dev.langchain4j.data.segment.TextSegment>> findRelevant(dev.langchain4j.data.embedding.Embedding referenceEmbedding,
 int maxResults,
 double minScore)

Specified by:

findRelevant in interface dev.langchain4j.store.embedding.EmbeddingStore<dev.langchain4j.data.segment.TextSegment>

fromAzureScoreToRelevanceScore

protected static double fromAzureScoreToRelevanceScore(double score)

Calculates LangChain4j's RelevanceScore from Azure AI Search's score.

Score in Azure AI Search is transformed into a cosine similarity as described here: https://learn.microsoft.com/en-us/azure/search/vector-search-ranking#scores-in-a-vector-search-results

RelevanceScore in LangChain4j is a derivative of cosine similarity, but it compresses it into 0..1 range (instead of -1..1) for ease of use.