The growing adoption of NoSQL document-oriented databases, such as MongoDB, has introduced new paradigms for managing data with flexibility and scalability. These characteristics make document models suitable for applications that involve heterogeneous data sources and evolving information needs. Current design methodologies often lack formal support, covering only a subset of Entity-Relationship (ER) constructs or relying on general-purpose notations like UML, which are not optimized for database design. In this work, we propose ER \(_{d}\) , a revised version of the ER model to support the conceptual design of document-oriented databases. ER \(_{d}\) extends the classical ER framework to capture the hierarchical and flexible nature of document structures, enabling a more accurate and semantically rich representation of data. We define a set of translation rules that map conceptual constructs into typical document patterns like embedding, referencing, and polymorphism.