Graph Studio FAQ

General Questions

What is Graph Studio?

Graph Studio is a fully managed, self-service graph data management and analytics environment. It is available as a feature of Oracle Autonomous Database Serverless (ADB-S) to store, manage, and analyze data as a graph.

How can I access Graph Studio?

Graph Studio is among the tools you can find under the Tool Configuration tab of your Autonomous Database dashboard. Alternatively, you can sign into Database Actions and click Graph Studio on the launchpad page under the Development group. To sign into Graph Studio, you need to create a database user with the GRAPH_DEVELOPER role assigned.

Graph Studio offers comprehensive functionality through a browser-based interface. For applications that require access to the functionality through a programming interface (API), there is a client-side library, which is included in the Graph Server and Client Kit, available for download here or on Oracle Cloud Marketplace.

What kind of analytics are possible with Graph Studio?

The graph features of Autonomous Database support both the property graph model and RDF graphs based on W3C semantic web standards such as SPARQL or OWL (W3C Web Ontology Language). There is even interoperability between the two graph models.

Using Graph Studio, you can perform pattern-matching queries using a SQL-like language, PGQL, which allows you to specify graph patterns following the language specification published here. You also have access to more than 60 graph algorithms for ranking and walking, community detection, path-finding, or evaluating graph structures, link prediction, or other purposes that can all be invoked using a Java or Python API. Leveraging machine learning on your property graph data in Graph Studio through the Graph Neural Network (GNN) algorithms in the PGX machine learning library, or through Oracle Machine Learning (OML), is possible, too.

The RDF Graph model is particularly well suited for knowledge graphs or application integration solutions. With RDF Graph, you are using Autonomous Database as a triple store. You can query and analyze your graph using SPARQL as the query language, or as a protocol, or you can perform inferencing based on ontologies defined with OWL.

Where can I find the documentation?

The documentation is included in the Autonomous Database documentation set here.

Is Graph Studio available on-premises?

No. Graph Studio is tightly integrated with Oracle Cloud Infrastructure (OCI), and there are currently no plans to make it available on-premises. For graph analysis on-premises or on Oracle Cloud IaaS, we recommend the Oracle Graph Server and Client, which is available for download here or on Oracle Cloud Marketplace, and works with any supported Oracle Database.

Pricing and licensing

What is the price of graph features in Autonomous Database Serverless?

When licensed with Autonomous Database, both Autonomous Data Warehouse and Autonomous Transaction Processing, there is no additional charge for Graph Studio or the graph features it is based on. There is, however, a usage-based cost. The Graph Studio environment uses additional ECPUs, which are charged at the rate of an ADB ECPU. You can set the number of ECPUs for Graph Studio in the Tools Configuration tab of the ADB Console page.

For more details on licensing of Autonomous Database, see the related FAQ.

Does Graph Studio offer free trial licenses?

Yes. Autonomous Database offers free 30-day trials, and Graph Studio is included. For more information, including how to sign up, click here.

Is Graph Studio included in a completely free service?

Yes. Autonomous Database Serverless offers a free service, and Graph Studio is part of this offering. For more information, including how to sign up, click here.

Security

Which roles and privileges are required to use Graph Studio?

As a database user, you can access the graph features once you have the GRAPH_DEVELOPER role assigned.

How does Graph Studio provide data security?

All the security mechanisms of Autonomous Database apply. For more information, including how to sign up, click here.

Architecture

What are the main architectural components of Graph Studio?

Graph Studio uses Autonomous Database as its persistence layer. From there, graph queries can be run in the database, or graph data can be moved into in-memory data structures for running graph algorithms and queries. You can use Graph Studio to manage the data, map relational tables to a graph model, and analyze the resulting graph data.

What are the main features of Graph Studio?

Graph Studio is an integrated, self-service tool to work with graphs. It includes:

  • A modeling tool to map relational data to a graph model.
  • A browser-based Jupyter Notebook environment enhanced with graph-specific features such as visualization, query, and analytics.
  • A graphical user interface to manage graph data.

It also includes sample data and demo notebooks to illustrate how to get started.

Loading data

Which tools can be used to load data into Graph Studio?

Graph Studio uses the underlying Autonomous Database for data storage, so any tools or solutions working with ADB as a database can be used. This includes Database Actions, GoldenGate, and Oracle Data Integration. For more details on loading data into the Autonomous Database click here.

Query and analysis

How can I search for graph patterns in my data?

For graph pattern matching, Oracle Autonomous Database offers support for the Property Graph Query Language (PGQL). PGQL is a SQL-based query language for the property graph data model. It allows you to specify high-level graph patterns that are matched against vertices and edges in a graph. Among other things, PGQL has powerful regular expression constructs for graph reachability, finding shortest paths, and finding cheapest paths. The latest language specification, along with many examples is available here. The new graph syntax available in the SQL:2023 standard published by ISO is available now in the OCI Oracle Base Database service and is planned to be available in Graph Studio in Autonomous Database.

Which graph algorithms does Graph Studio support?

Graph Studio includes more than 60 optimized, built-in graph algorithms for tasks such as ranking and walking, community detection, pathfinding, link prediction, and structure evaluation. In addition, we have a machine learning library including algorithms such as DeepWalk and supervised and unsupervised GraphWise. For more details on the individual algorithms click here and here for the PGX.ML library, respectively.

Which programming languages are supported in Graph Studio?

Currently, the Notebook in Graph Studio includes a Java interpreter and a Python interpreter to access the Graph APIs and a PGQL interpreter to execute pattern-matching queries against the graph. In addition, there is a SQL interpreter to access data in relational tables, as well as a SPARQL interpreter to work with RDF Graphs. The new graph syntax available in the SQL:2023 standard is planned.

Can I use Oracle APEX as a low-code application development environment for graphs?

Yes, with the APEX Graph Visualization plug-in, you can seamlessly work with the graphs you have created in Graph Studio. Specifically, you can build applications that visualize these graphs and offer interactive analysis to the user.

Performance

How can I control the amount of memory dedicated to graph analytics?

Graph Studio includes a calculator to estimate how much memory is required to move the graph data into in-memory data structures for analysis. To speed up the process, it is possible to pre-allocate memory in Graph Studio.

Migration

How can I use a PGQL DDL statement I have created previously to map relational data to a graph model?

The Graph Modeler in Graph Studio generates PGQL DDL as part of the modeling process. If you have already created a PGQL DDL statement previously, you can copy and paste it into the Graph Modeler and either work with it as-is or refine it if needed. Alternatively, you can enter the PGQL DDL statement using the '</> Query' button under the Graphs menu icon.