In Fusion Data Intelligence, per-dimension data security rules are defined within the semantic model extensions in the FDI console. This layer describes dimensions, measures, and how data is sliced, enabling value-specific filters. The Admin Console handles broader access, while higher-level security policies govern authentication.

Multiple Choice

Where can you implement custom data security rules that apply to specific dimension values in FDI?

In Fusion Data Intelligence, custom data security rules that target specific dimension values are defined where you model the data—the semantic model extensions in the FDI console. The semantic model is the layer that describes dimensions, measures, and how data can be sliced and filtered. By adding data security rules there, you can specify filters that apply to particular dimension values (for example, limiting data to certain product categories or regions). These rules are evaluated as part of the semantic model’s logic, ensuring users only see data for the dimension values they’re permitted to access. Other locations aren’t designed for this kind of per-dimension, value-specific control. The Admin Console security page handles broader admin access, not granular dimension-value restrictions. Fusion Applications security policy and Oracle Identity Cloud Service govern higher-level authentication/authorization and enterprise security, not the intra-model, dimension-level data filters used by FDI.

Data never exists in a vacuum. It travels through dashboards, models, and policies, picking up the shape of who can see what along the way. When you’re working with Fusion Data Intelligence (FDI), you’re not just organizing data—you’re encoding the rules that decide which slices of data are accessible for each user. And the most precise way to govern that access, down to specific dimension values, lives right in the semantic model extensions in the FDI console. Here’s why that matters, and how to think about it in practical terms.

A quick hello to the semantic model: where the data really meets the rules

Imagine you’re shaping a data cloud that needs to be navigated by several teams: sales, marketing, finance, and product ops. Each group has its own view of the data, not just in terms of which tables or facts they can reach, but in how those facts get sliced. The semantic model is the layer that defines dimensions (like product category, region, or time), measures (such as revenue or units sold), and the rules that govern how data can be filtered or aggregated. It sits between the raw data and the end-user visuals, acting like the map and the compass rolled into one.

Now, put data security rules into this layer, not later in the chain. When you add data security rules at the semantic model extensions level, you’re embedding access constraints where decisions are actually made as users interact with the data. It’s a bit like setting guardrails directly on the highway ramps—filters that say, “If you’re looking at the product category ‘Electronics,’ you’ll only see regions where you’re permitted to operate,” or “For the dimension ‘Department,’ you’ll only view the values you’re allowed to slice through.” This is granular control, and it travels with the data, not as an external toll gate that could be bypassed or misapplied.

Why per-dimension, value-specific rules are different from generic access controls

It’s tempting to think access control is just “who can log in” or “who can see the dashboard.” Those broad controls are essential, of course—they sit in the higher layers of security infrastructure. But the real nuance lives closer to the data itself. Per-dimension rules let you answer questions like:

  • Can a regional sales manager see revenue for their country only, not the entire region?

  • Are product managers allowed to compare performance across all product families, or should they be restricted to their own product lines?

  • Should a finance user see budgets only for the business units they’re responsible for, even when the same data source feeds multiple departments?

These questions aren’t about authentication; they’re about the scope of visibility within the data model. By defining them in the semantic model extensions, you ensure the filters are an inherent part of how data is ingested, modeled, and presented. That leads to consistent results across dashboards, reports, and explorations—without needing to chase people to apply the right filters manually every time.

How it works under the hood

The semantic model extensions are, essentially, the blueprint for how data should be interpreted by end users. When you introduce security rules into this blueprint, you’re attaching logic to dimension values. The rules specify which dimension values are accessible to which user groups or roles, and they’re evaluated as part of the model’s logic. If a user drills into a view that contains a restricted dimension, the underlying filter activates automatically, hiding disallowed values and showing only what’s permitted.

The beauty of this approach is consistency. If a rule says “limit by region to North America for the Global Sales team,” that constraint applies across all visualizations built on the semantic model—no exceptions, no manual toggling. The result is a coherent, auditable data experience.

Practical examples to illustrate the approach

  • Region-based access: A regional manager should only see data for the regions they’re accountable for. You could set a rule on the Region dimension that maps the manager’s role to a subset of values like North America and Europe, excluding Asia-Pacific.

  • Category-specific limits: A product manager might be allowed to view performance only for their product categories. The Extension lets you tie a user or role to specific dimension values like “Wearables” or “Home Electronics,” filtering out everything else.

  • Time-bound visibility: For a subset of users, you might want to see data for the current quarter only, while others can view the full fiscal year. Time constraints can be woven into the semantic model so that the date dimension enforces the appropriate window.

Note that these rules live where the data is defined, not in external policy documents that need separate maintenance. That makes governance more straightforward and less brittle.

What you don’t use for this level of granularity

It’s useful to separate the different layers of security so you don’t end up with a tangled web. The Admin Console security page is invaluable for administering who has access to the system itself (logins, roles, and broad admin permissions). It’s about governance at the IT and admin level, not about per-dimension data restrictions. Likewise, higher-level policy frameworks—like Fusion Applications security policy or Oracle Identity Cloud Service—exist to control authentication, role assignment, and enterprise-wide security posture. They aren’t the right tools for enforcing granular data filters inside a semantic model.

So, if your goal is fine-grained, value-specific control within a data model, you’ll want to stay focused on the semantic model extensions in the FDI console. That’s where the rules become part of the data’s natural behavior rather than something tacked on at the edges.

A few best practices to keep in mind

  • Start with a clear map of dimensions that matter for security. The obvious suspects are Region, Product Category, and Time, but also look at customer segments, channels, or business units. The more complete your dimension map, the easier it is to describe who can see what.

  • Use roles that align with real-world responsibilities. Instead of a monolithic “analyst” role, consider roles like Regional Analyst, Product Manager, or Finance Controller, and tie those roles to specific dimension values.

  • Keep rules readable and maintainable. A rule that says “Region in ('NA', 'EU') and Category in ('Electronics', 'Appliances')" reads better than a tangled jungle of nested conditions. Document the intent as you implement it, so future you or teammates don’t have to reverse-engineer the logic.

  • Favor incremental, testable changes. Add a small rule, verify it across multiple dashboards, then expand. This helps catch edge cases where a dimension value behaves differently than expected.

  • Audit and monitor. Build in checks that verify which users can access which data slices. Regular audits help catch drift if roles evolve or data landscapes shift.

From theory to everyday use

If you’re hands-on with FDI, you’ll find the semantic model extensions to be a powerful place to encode data governance right where data gets interpreted. It’s a design choice that translates into practical benefits: fewer permissions bottlenecks, more predictable data behavior, and a governance story that travels with the data—no matter which report or analysis a user opens.

A moment to breathe with a digression—data governance is not a lecture hall drill

People often imagine governance as a heavy, bureaucratic overlay. In reality, when you configure these rules in the semantic model, you’re doing something almost polite: you’re guiding the data’s journey with gentle, precise checks that respect user roles while preserving the data’s integrity. It’s a quiet power, the kind that makes dashboards feel trustworthy and reports feel reliable without requiring you to babysit every click.

What to watch out for as you implement

  • Over-filtering can backfire. If you tether rules too aggressively, you risk rendering users with too little data to be meaningful. Balance security with usability by validating that every role has a productive view.

  • Dimension value naming matters. If dimension values drift or get renamed, your rules might silently constrain the wrong things. Establish a governance rhythm for maintaining dimension value master data alongside your rules.

  • Cross-dimensions interactions can surprise you. A filter on Region might interact with a filter on Channel in unexpected ways. Test end-to-end scenarios to catch these interactions early.

Putting it all together: why the semantic model extensions matter

The essence is straightforward: you want data to reveal itself only in the ways you’ve agreed. The semantic model extensions in the FDI console provide a precise, centralized way to encode those permissions at the level where data is defined and interpreted. It’s where the meaning of the data is set, and with it, the meaning of your security rules.

If you’re curious about how your team can bring this into daily practice, start by inventorying the dimensions that matter for governance, align them with the roles in your organization, and then craft a small, well-documented rule set in the semantic model. See how the data’s behavior improves across a few representative dashboards. If it feels right, you’ve probably found a sweet spot where governance and usability meet—without fuss.

In the end, data security isn’t about building walls so high you can’t see over them. It’s about shaping a navigable landscape where the right people can explore the right data, in the right way. The semantic model extensions in the FDI console are the compass that helps you do precisely that—mapping permission to dimension values so the data speaks clearly to every user who deserves to see it. And that clarity? It’s what makes data-driven work not just possible, but genuinely empowering.