Integrated Graph Database
Graphs enable you to find connections and explore relationships in your data. Oracle Graph is an AI-ready, integrated feature of Oracle's converged database that eliminates the need for a separate graph database and data movement. Analysts and developers can address various use cases, including financial fraud detection and manufacturing traceability, while gaining enterprise-grade security, ease of data ingestion, and strong support for operational workloads.
Annuncio sullo sviluppo generativo (GenDev) per le aziende
Guarda la replica della keynote dell'EVP Juan Loaiza all'Oracle CloudWorld per scoprire di più su questa rivoluzionaria infrastruttura AppDev incentrata sull'intelligenza artificiale.
Why use Oracle Graph?
-
Find the unexpected
Data is connected. Discover hidden patterns and find new insights easily and quickly with more than 80 prebuilt algorithms, automated analysis, visualization tools, and graph-powered AI using RDF or property graphs.
-
Real-time graph analytics
Make informed business decisions with graph analytics based on operational and transactional data in your Oracle Database.
-
Enterprise-grade
Benefit from scalability, high availability, security, AI capabilities, and other converged features of Oracle Database when running graph analytics.
Graph use cases
Security
Security is a must-have for today’s organizations. Graphs can help you quickly pinpoint connections within complex networked data, including interlinked IT systems and criminal networks, so your organization can better respond to threats and identify bad actors.
-
Cybersecurity
Graphs are ideal tools for cybersecurity because they capture and model activities and events in complex networks within the IT infrastructure. You can combine graph analysis with machine learning (ML) to quickly uncover connections, patterns, and anomalies within large volumes of data and provide an interactive, visual way to explore security data to detect threats, invalid traffic, and malware. Automating graph analysis in threat intelligence saves time and employee labor versus manual investigations.
-
Criminal investigation
Crimes don't often happen in silos. Frequently, there are many interconnected people, organizations, and locations. Putting data into graphs enables law enforcement to efficiently identify criminal networks and spot patterns.
-
Data privacy and GDPR
Finding out which user has access to a specific data element can be a tedious task if multiple layers of software are involved. Graphs help to keep track of these kinds of indirect relationships and enable organizations to prove regulatory compliance easily.
AI and machine learning
AI and machine learning are important new technologies because of their promise to improve business results and create new impacts. Graphs are useful in improving the accuracy of predictions from ML models as they offer a complementary view on the data.
-
GraphRAG
Take your generative AI–powered similarity searches to the next level with GraphRAG. By integrating relevant business data, retrieval-augmented generation (RAG) can enhance user queries through LLMs. GraphRAG can improve the accuracy and relevancy of the responses by leveraging graphs that capture and provide rich insights into relationships between data entities, such as social networks, financial flows, or supply chains. It also allows you to trace back the responses to their origins to prove the correctness for compliance or validation.
-
Feature engineering
Some feature engineering tasks are complicated to accomplish, and graphs can help simplify these tasks. For example, taking into account indirect relationships between entities or determining clusters of closely connected entities can be cumbersome without using graphs. Running graph algorithms on a data set creates enriched data which can then be used for machine learning models as features.
-
Graph neural networks
The use of graphs as a recommendation engine is well known, but graphs can also be used for predictive recommendations. For example, an online retail store wants to send recommendations to a customer, with timing determined by when the customer is predicted to run out of the item. Graph neural networks, which can capture the graph itself as an input of machine learning and neural networks, provide potentially better accuracy because the graph can hold more information than relational tables.
Government
Governments can use graph technologies for defense and public safety, to aid in public health initiatives, and for linked open data initiatives for their citizens.
-
Tax fraud
Resource-strapped governments deal with inventive and evasive criminals. Graphs can help organizations understand the structure of shell corporate entities, improve manual investigation with visualization tools, and discover suspicious patterns to trace the paths in complex networks that ultimately lead back to the defrauder.
-
Criminal investigation
Large-scale crimes frequently involve many interconnected people, organizations, and locations. Putting data into graphs enables law enforcement to efficiently identify criminal networks and spot patterns.
-
Contact tracing
Disease contact tracing has been a time-critical, urgent activity worldwide. Graphs are ideal for analyzing disease patterns. Analysts can use the information on the people who have tested ill—their interactions with others and visited places—to help rapidly locate hotspots and connections to prevent further outbreaks.
Manufacturing
Turning raw materials into products involves many relationships, components, and dependencies, which makes graph technologies a perfect fit for speedily discovering more information.
-
Bill of materials
A product may have tens of thousands of parts. What if you need to quickly find the impact of changing one part—or a few parts? What if each part has multiple dependencies? Graph analysis enables real-time interactive analysis of such queries.
-
Master data management
At many manufacturing factories, each department may use a different name for the same component. Problems arise when you need to find out more about certain use cases and which components are involved for that specific item. RDF graphs enable you to model different components and use the relationships and connections they have with each other.
-
Traceability
Traceability is important in situations such as product recalls when you need to trace a specific component that was produced from a specific factory during certain dates and times. Identifying cars or other products in the market by tracing back components can be very difficult without graph technology.
Marketing
To appeal to their target audience, marketers must understand their customers and their relationships with their products.
-
Customer 360-degree analysis
Today, companies know more about customers through master data, transactions, bid data, predictions, but they often don’t leverage this information fully. Creating a true customer 360 analysis is difficult even when the data is collected and integrated into the physical platform. Graphs can logically integrate the data and simplify creating a unified view of each customer.
-
Product recommendations
While non-graph technologies can support recommendations, graphs enable higher accuracy because they can add context. Graph databases emphasize connections such as the relationships between customers and the products they like to buy providing more contextual input to the recommendation process.
-
Social media
Social media is driven by relationships, connecting users throughout the world. Ensuring the validity of those users is key. Graph can traverse social networks and related data very quickly, providing recommendations of users, images, products, while also detecting fraudulent activity and sock puppet accounts.
Financial services
No matter how hard they try to disguise it, financial criminals are linked by relationships—to other criminals, locations, or bank accounts. Graph technology takes advantage of this fact to unfold new possibilities to fight criminals.
-
Money laundering
Criminals try to hide fraudulently obtained money through a long and complex series of valid transfers between legitimate accounts. Graphs make it easier to detect fraud by analyzing transactions between entities and identifying those with similar information, revealing accounts who are sending money to each other.
-
Detecting money mules and mule fraud
Traditionally, alerts from rule-based models combined with manual inspection are used to detect money mules and mule fraud. Machine learning is also used to predict human decisions. However, improving such models is difficult due to the accounts’ limited information. Graphs go beyond this limitation by taking the transaction information as edges and generating more features of the accounts based on surrounding relationships and transactions.
-
Real-time fraud detection
Consumers demand instant access to services and money transfers, creating opportunities for criminals. Because graphs enable lightning-fast answers to queries and expanded access to data, they have become a popular technology in real-time fraud detection. Property graph is often used, especially in online banking and ATM location analysis, because graphs help enhance fraud detection algorithms using data that's otherwise difficult to associate.
Oracle Graph pricing
Graph database and graph analytics are integrated into Oracle Database and included in on-premises database licenses and Autonomous Database.
See how Oracle's graph database makes it easy to explore relationships and discover connections in data by providing support for different graph structures, powerful analytics, and intuitive visualization.
Learn how to create property graph models to analyze complex relationships using simple SQL without data duplication. Discover how to use knowledge graphs to boost machine learning models and retrieval-augmented generation (RAG) for AI-powered apps.
Create and manage SQL property graph models with a few clicks—without manually coding SQL syntax. Explore complex connections in data without duplicating the data.
Knowledge graphs, also known as ontologies, help applications query data with the associated context and enable users to make business decisions based on the context. Learn how Oracle Graph supports such ontologies with a utility use case.
Bring the Power of Graphs to Generative AI
Melliyal Annamalai, Distinguished Product Manager, Oracle
Generative AI can answer questions in human language on a wide range of topics using large language models (LLMs) that are trained on available data. But how about data the LLM has not been trained on, such as your own business data? How can we use the power of generative AI on such data? We have the answer—Graph RAG. And Oracle has made harnessing the power of Graph RAG much easier.
Read the complete post
