In part 8 of the Semantic Model Series, I show how to implement dynamic row-level security in Oracle Analytics using data filters at the Business Model level. I walk through setting up users, groups, and application roles, then applying data filters tied to session variables so access to regions like Americas and EMEA is enforced automatically — no per-report filtering needed.
In part 7 of the Semantic Model Series, I cover the three types of variables in Oracle Analytics Semantic Modeler — session, global, and static — and how each is created through initialization blocks. I show practical examples of testing each variable type in analyses, and set the stage for using session variables to implement row-level security at the model level.
In part 6 of the Semantic Model Series, I cover how to build calculations in Oracle Analytics' logical layer — from simple measures like Profit and Profit Margin %, to Time Series calculations (AGO, TODATE) and Level-Based Measures for market-share style analysis. I also walk through reorganizing the Presentation Layer into logical folders and testing everything after deployment.
In part 5 of the Semantic Model Series, I cover the Presentation Layer in Oracle Analytics Semantic Modeler — subject areas, presentation tables, columns, and hierarchies that shape what end users see. I walk through creating a subject area, dragging in logical tables, running a consistency check, deploying the model, and testing it in a live workbook.
In part 4 of the Semantic Model Series, I dive into the Logical (Business) Layer in Oracle Analytics — how it abstracts physical data into business-friendly facts and dimensions. I walk through building fact and dimension tables, creating logical joins and hierarchies, and configuring time dimensions with chronological keys so time-series functions like AGO and TODATE work correctly.
In part 3 of the Semantic Model Series, I cover the Physical Layer in Oracle Analytics — the foundation that connects to source databases and defines tables, aliases, and joins. I walk through creating connection pools, importing fact and dimension tables, building a custom table from a SQL query, and setting up table aliases and joins to prepare the model for the Logical Layer.
In part 2 of the Semantic Model Series, I walk through creating a new semantic model in Oracle Analytics — choosing between an empty model, RPD import, deployed-model import, or Git clone — and reviewing the six main tabs: Connections, Physical Layer, Logical Layer, Presentation Layer, Variables, and Invalid Files. I also cover setting up the connection used to build out the model in later parts.
In part 1 of the Semantic Model Series, I explain the difference between Regular Connections (for data retrieval and analysis) and System Connections (required for building semantic models) in Oracle Analytics. I walk through creating both connection types step by step, including a tip on keeping connection names short for better compatibility with the Semantic Modeler.