개요
Native graph experience
Spanner Graph supports ISO Graph Query Language (GQL), the international standards for graph databases. It offers an intuitive and concise way to match patterns, traverse relationships, and filter results in graph data, making it easier to uncover hidden relationships and insights.
Unified relational and graph models
Spanner Graph bridges the relational and graph worlds. By combining the strengths of SQL and GQL, it enables analysts and developers to query structured and connected data in a single operation. Additionally, Spanner Graph simplifies graph creation with a declarative schema, transforming your relational data into rich, interconnected graphs, ready for exploration and analysis.
Built-in search capabilities
The integrated full-text search and vector search capabilities in Spanner Graph enable you to deliver a new class of AI-enabled applications. You can search for relevant nodes and edges based on either semantic similarity using vector search or specific keywords using full-text search respectively, then explore the rich context surrounding these elements using graph.
Industry-leading scalability, availability, and consistency
Built on Spanner’s unmatched scale, 99.999% availability, and consistency, Spanner Graph is an ideal choice even for mission-critical applications. Built-in sharding automatically distributes data for horizontal scalability and helps ensure uninterrupted access to critical graph data. By maintaining a consistent view of nodes and edges across the entire graph, it eliminates the risk of dangling edges or other inconsistencies.
AI-powered insights
Spanner Graph is integrated with Vertex AI. You can directly access Vertex AI's extensive suite of predictive and generative models through the Spanner Graph schema and query, which streamline your AI workflow.
Enterprise-ready
Spanner Graph inherits all the enterprise readiness from Spanner, offering customer-managed encryption keys (CMEK), data-layer encryption, IAM integration for access and controls, and comprehensive audit logging. In addition, it supports VPC-SC, Access Transparency, and Access Approval. Furthermore, fine-grained access control lets you authorize access to Spanner data at the table and column level.
작동 방식
To create a graph, first create tables in Spanner to store entities and relationships, then map the tables to a graph using the graph schema. You can use GQL to query the graph, or combine GQL with SQL to query the graph and tables together.
To create a graph, first create tables in Spanner to store entities and relationships, then map the tables to a graph using the graph schema. You can use GQL to query the graph, or combine GQL with SQL to query the graph and tables together.
일반적인 용도
Knowledge graph
With Spanner Graph, you can develop knowledge graphs that capture the complex connections between entities, represented as nodes, and their relationships, represented as edges. These connections provide rich context, making knowledge graphs invaluable for developing knowledge base systems and recommendation engines. With integrated search capabilities, you can seamlessly blend semantic understanding, keyword-based retrieval, and graph for comprehensive results.
방법
With Spanner Graph, you can develop knowledge graphs that capture the complex connections between entities, represented as nodes, and their relationships, represented as edges. These connections provide rich context, making knowledge graphs invaluable for developing knowledge base systems and recommendation engines. With integrated search capabilities, you can seamlessly blend semantic understanding, keyword-based retrieval, and graph for comprehensive results.
Recommendation engine
Spanner Graph naturally models relationships between entities (like users, products, or friends), making it simple to traverse connections and uncover potential relationships. Built-in vector search and full-text search enable similarity-based recommendations based on product user profile, product descriptions, and reviews. This combination enables highly relevant and personalized recommendations across diverse domains, all within Spanner Graph.
방법
Spanner Graph naturally models relationships between entities (like users, products, or friends), making it simple to traverse connections and uncover potential relationships. Built-in vector search and full-text search enable similarity-based recommendations based on product user profile, product descriptions, and reviews. This combination enables highly relevant and personalized recommendations across diverse domains, all within Spanner Graph.
Financial fraud detection
Spanner Graph naturally models the complex relationships between financial entities like accounts, transactions, and individuals, making it easier to identify suspicious patterns and connections that could signal fraudulent activity. In addition, built-in vector search reveals hidden connections and anomalies in the embedding space. By combining these technologies, financial institutions can create comprehensive fraud detection systems that can quickly and accurately identify potential threats, minimizing losses.
방법
Spanner Graph naturally models the complex relationships between financial entities like accounts, transactions, and individuals, making it easier to identify suspicious patterns and connections that could signal fraudulent activity. In addition, built-in vector search reveals hidden connections and anomalies in the embedding space. By combining these technologies, financial institutions can create comprehensive fraud detection systems that can quickly and accurately identify potential threats, minimizing losses.
GraphRAG
Spanner Graph takes Retrieval Augmented Generation (RAG) to the next level by grounding foundation models with rich contextual information stored in a knowledge graph. Where traditional RAG provides LLMs with chunks of data extracted from a source document, GraphRAG goes a step further by including the relationships with other content to facilitate a more comprehensive understanding and inference.
방법
Spanner Graph takes Retrieval Augmented Generation (RAG) to the next level by grounding foundation models with rich contextual information stored in a knowledge graph. Where traditional RAG provides LLMs with chunks of data extracted from a source document, GraphRAG goes a step further by including the relationships with other content to facilitate a more comprehensive understanding and inference.
Gaming
Game worlds can be represented as entities like players, characters, items, and locations as nodes, and relationships between them as edges. This structure enables efficient traversal of connections, essential for game mechanics like pathfinding, inventory management, and social interactions. Spanner Graph’s scalability ensures that the database can handle the influx of players during peak times or major events, preventing lag and server crashes that would disrupt gameplay and frustrate users.
방법
Game worlds can be represented as entities like players, characters, items, and locations as nodes, and relationships between them as edges. This structure enables efficient traversal of connections, essential for game mechanics like pathfinding, inventory management, and social interactions. Spanner Graph’s scalability ensures that the database can handle the influx of players during peak times or major events, preventing lag and server crashes that would disrupt gameplay and frustrate users.
Get started with Spanner Graph
Learn how to use Spanner Graph
Create a 90-day Spanner instance for free
Introducing Spanner Graph: Graph databases reimagined
Explore graph capabilities
Learn more about Spanner
비즈니스 사례
"At Credit Karma, ensuring the safety of our 130+ million members’ data is our top priority. To combat and eliminate fraud across our systems, we’ve partnered with Google to enhance our fraud mitigation capabilities by implementing Google Spanner Graph Database. This advanced platform capability allows us to detect potential fraud threats before they happen. With Spanner Graph, we effectively detect and prevent fraudulent transactions, account takeovers, and other fraudulent activities. " - Credit Karma
Contact usFeatured benefits
Elevate your gen AI applications with knowledge graph-powered intelligence.
Discover hidden connections and relationships in your data.
Streamline operations with a single unified database, integrating relational, graph, search, and key-value capabilities.
Social networks
Individuals, groups, interests, and interactions can be represented as nodes and edges for efficient analysis of connections and the discovery of patterns such as mutual friends, shared interests, or overlapping group memberships. These insights can then be leveraged to generate personalized friend and content recommendations as well as ad targeting. Additionally, integrated full-text search allows you to easily find people, groups, posts, or specific topics using natural language queries.
방법
Individuals, groups, interests, and interactions can be represented as nodes and edges for efficient analysis of connections and the discovery of patterns such as mutual friends, shared interests, or overlapping group memberships. These insights can then be leveraged to generate personalized friend and content recommendations as well as ad targeting. Additionally, integrated full-text search allows you to easily find people, groups, posts, or specific topics using natural language queries.