The article presents a formal way to model a multilevel security system for Object-Oriented Databases. The researchers use a language based on logic to represent database content and integrity constraints. They start with a basic security model called Single-View, where each piece of information is classified based on its level of sensitivity. By refining this model, they develop the MultiView model, which includes new security features like protecting the database schema. The researchers also provide general theorems that must be followed when classifying information. Overall, the study aims to enhance the security of Object-Oriented Databases by extending the existing MultiView model.