Modeling

• consists of building an abstraction of reality

• model

  • abstract description of a problem/problem solution
  • abstractions are simplifications because
    • they ignore irrelevant details
    • they only represent the relevant details

  models must be falsifiable

  • testing: the act of disproving a model
    • any model is a theory
    • build models and try to find counter examples by requirements validation, user interface testing, review of the design, source code testing, system testing, etc.

• we use it to produce software because

  • software is getting increasingly more complex
  • modelling is a means for dealing with complexity

• key technique: abstraction

  • is a thought process where ideas are distanced from objects
  • allows us to ignore unessential details
  • as activity ⇒ is the resulting idea of a thought process where an idea has been distanced from an object
  • as entity ⇒ ideas can be expressed by models

UML (Unified Modelling Language)

• supports the 3 ways to deal with complexity

  • Abstraction and Modeling
  • Decomposition
  • Hierarchy

• nonproprietary standard for modeling software systems

• you can solve 80% of the modeling problems by using 20 % UML

  • 80-20 rule: Pareto principle

    

• is specified in a more rigorous manner by means of concrete syntax, abstract syntax and semantics

• abstract syntax is described by means of the UML metamodel and WFRs (invariants on metamodel classes)

• language describing the UML metamodel, MOF is a subset of the UML language

• beginning with the first version, UML included the OCL

UML is a language supporting model specification

• model
  • is an abstraction describing a problem/problem domain, problem solution/ system or a subsystem:
    • a system that no longer exists,
    • an existing system,
    • a future system to be built
• view
  • depicts selected aspects of a model
• diagram/notation
  • a set of graphical/textual rules for depicting views and formal specifications