By Srivyshnavi Akarapu, HEXstream data analytics engineer

In this blog, we will go through the semantic model in Oracle Fusion Data Intelligence (FDI) and understand the available options in the semantic model. We will also see how these options help with customizing the model and getting data ready for reporting.

In today's data-driven world, organizations rely on dependable, consistent and useful data insights to make better decisions. Oracle Fusion Data Intelligence plays a key role in transforming raw data into meaningful insights.

FDI is an advanced analytics platform that integrates data from Oracle Fusion Applications and other sources into a unified environment. It provides prebuilt dashboards, KPIs and data models to help organizations make faster and smarter decisions.

The semantic model is the key component in this process. It enhances overall reporting and analysis and makes data more useful and easier to understand, particularly for business users.

Semantic model

A semantic model in FDI is a logical layer that organizes data into business terms such as measurements and dimensions, enabling users to perform analysis and create reports without advanced technical understanding.

We have different options to customize and manage the semantic model, including:

• Sandbox
• Create Logical Star 
• Edit Logical Star 
• Manage Dimensions 
• Manage Extensions 
• Create Subject Area 
• Modify Subject Area 
• Validate Sandbox 
• Apply Changes 
• Merge Sandbox to Main 
• Publish Model
• Validate Model 
• Manage Variable
• Create Variable
• Edit Variable

Let's explore further:

Sandbox—In the semantic model, the sandbox serves as the primary workspace for creating and validating customizations. To start customization, a sandbox is created in the development or testing environment. This helps you make changes safely without impacting the main application. Once the customizations are completed and validated, the required objects can be migrated to other environments using the Bundle feature.

Create Logical Star—In FDI, a Logical Star helps organize data in manner that eases reporting and analysis. It connects fact tables with dimension tables. This structure makes the data clearer for business users to understand. You can create a custom logical star when you add new or additional data to the warehouse, which is particularly useful when you require new reports based on business requirements or when existing models are not sufficient for new requirements. By creating a logical star, you can:

• Combine custom fact and dimension tables. 
• Define how tables are connected (joins) 
• Create new calculations and metrics
• Prepare data in a structured way for reporting.

Edit Logical Star—In the semantic model, we have an option to edit a logical star. We can work with both custom logical stars and OOTB logical stars, but there is a difference. With Custom Logical Star → We can fully edit it (add or remove columns, joins, calculations). With OOTB Logical Star → We cannot change it directly, but we can extend it by adding new columns or joins.

Manage Dimensions—In the semantic model of FDI you can create custom dimensions based on your business needs. You can connect these dimensions to existing or custom fact tables. After that, you can add them to subject areas so they can be used in reports.

Manage Extensions—Using Manage Extensions we can modify both OOTB and custom dimensions. In this step, you can:

• Extend existing dimensions
• Add hierarchies
• Add new columns

This helps to make your data more useful and better for reporting.

Create Subject Area—We can create a new subject area for reporting in the semantic model. A subject area works like a container. We can add facts and dimensions into it. We can also create a new subject area using the existing one. It helps to keep all data in one place, so it is easy to use in reports.

Modify Subject Area—In the semantic model, we can modify both custom and prebuilt subject areas. Custom Subject Area → We can change existing data and add new data elements. Prebuilt Subject Area → We can add more data elements.

Validate Sandbox—In the semantic model, we can validate the sandbox to check for errors before applying changes. When we run validation, it shows a list of errors if any are present. We need to review these errors and fix them in the sandbox.

Apply Changes—After creating the customization, we need to select the Apply Changes option in the sandbox. This step helps to validate the changes before merging with main or publishing the sandbox.

Merge Customization Sandbox to Main Sandbox—After preparing, validating and applying the changes in the sandbox, we can merge the sandbox with the main sandbox. This step moves all the latest changes from the sandbox to the main.

Publish Model—After merging the sandbox, we can publish the model in development or test environments. Publishing helps to make sure there are no errors and the changes are working correctly. While publishing, we can select the extensions and security settings that we added during customization. These will be applied to the model during publishing.

Validate Model—In the semantic model, we have an option called Validate Model. This option checks the entire model and gives the status if there are any errors. Before creating a bundle and deploying it to another instance, we use this option to detect and fix errors.

Manage Variables—In the semantic model, we can use Manage Variables to control how sessions and queries work. We can create and modify custom variables based on our requirements.

Create Variable—We can create custom session variables and use them in the semantic model. These variables are available for use after merging the sandbox into the main sandbox.

Edit Variable—We can edit or modify a variable to update its logic. This includes updating the SQL query that runs in the data warehouse and returns values that can be used in reports.

Conclusion: In FDI, the semantic model helps to organize data in a simple and straight-forward way for reporting. By using options like sandbox, logical star, dimensions, subject areas and variables, we can customize the model based on business requirements.

Each step—from creating and editing to validating, merging and publishing—helps in building a clear and error-free model. Overall, the semantic model makes reporting easier, more organized, and useful for business users.

CLICK HERE TO CONNECT WITH US ABOUT SEMANTIC MODELING.