The method and algorithms to find essential attributes and objects of Subject Domains

The paper proposes a new model, method and algorithms for processing a Subject Domain (SD). It considers an SD as a tuple (E,V,T), where E is a set of all objects of this domain, V is a set of all connections between these objects, T is a set of mass problems that can be solved in this domain. Each object is represented by its essential attributes. This allows us to describe SD, to solve most of the mass problems from T, and to control a state of the domain. The goal of the paper is to develop a method and algorithms that allow us to find essential attributes of domains, described as relational databases or production systems. For a production system, we build a dependency graph with attributes as vertexes, and calculate for each of the vertexes a value that indicates how many other attributes can be found using a particular vertex. High-ranked vertexes are considered as essential. The method is based on considering as essential of the most frequently accessed attributes in a relational database. Having essential attributes of objects of a domain, it allows us to define essential attributes of the domain itself.