Seamless Formalizing the UML Semantics through Metamodels

Despite the fact that the Unified Modeling Language (UML) has been adopted by the Object Management Group (OMG2 ) as the standard notation for use in Object-Oriented (OO) Systems Development, it still does not have a truly formal semantics. There is currently much effort directed towards formalizing particular aspects or models of UML. However, the literature gives little insight into the appropriate strategy for tackling this problem within an integrated basis including the language evolution. This chapter identifies and discusses three feasible strategies which can be applied to formalize UML. One of these strategies is selected to underpin the four-layer architecture on which UML is based. The approach is based on the soundness of algebraic specification theory, which, in addition, provides suitable theorem-proving capabilities for exploiting the UML formal model obtained. The formal models proposed are specified using an executable algebraic specification language called Maude.

Linking UML with Integrated Formal Techniques

The challenge for complex systems specification is how to visually and precisely capture static, dynamic and real-time system properties in a highly structured way. In particular, requirement specifications for composite systems often involve capturing concurrent interactions between software control parts and physical system components/devices. The requirement specifications of such systems need to capture the structure and behavior of each individual physical/software components and their communications. In this chapter, we investigate the links between the graphical notation UML and an integrated formal notation. We present an effective combination of UML and an integrated formal method for the requirement specification of a light control system. This work is supported in part by the research grant (Integrated Formal Methods) from National University of Singapore (No. RP3991615).

Extension of the Unified Modeling Language for Mobile Agents

Mobile agents gained immense attraction as a new programming concept for implementing distributed applications. However, up to now mobile agent programming has been mainly technology driven, with a focus on the implementation of mobile agent platforms and only small programming applications. In this chapter, we present an extension of the standard UML that provides language concepts for modeling mobility both in analysis and design phases. This extended version of UML is applied to the modeling of an advanced telecommunication system.

An Interactive Viewpoint on the Role of UML

The role of the Unified Modeling Language (UML) is to model interactive systems, whose behaviors emerge from the interaction of their components with each other and with the environment. Unlike traditional (algorithmic) computation, interactive computation involves infinite and dynamic (late binding) input/output streams. Tools and models limited to an algorithmic paradigm do not suffice to express and manage the behavior of today’s interactive systems, which are capable of self-reconfiguring and adapting to their environment. Whereas procedural languages may express precise designs of closed processes, UML’s objective is to provide support for the analysis and specification of increasingly complex and inherently open systems. Interactive systems require dynamic models where interaction has first-class status, and where the environment is modeled explicitly, as a set of actors whose roles constrain the input patterns through use cases. UML’s interaction-based approach to system modeling fits well with the encapsulation-based, object-oriented approach to implementation. By coupling these approaches, the software engineering process can promise to provide a more complete solution to system design and implementation, leading the way for widespread adoption of networked and embedded intelligent agent technology. A theoretical framework for modeling interactive computing can strengthen the foundations of UML and guide its evolution.

Using a Semiotic Framework to Evaluate UML for the Development of Models of High Quality

Many researchers have evaluated different parts of UML and have come up with suggestions for improvements to different parts of the language. This chapter looks at UML (version 1.3) as a whole, and contains an overview evaluation of UML and how it is supported in the modeling tool Rational Rose as a basis for creating models of high quality. The evaluation is done using a general framework for understanding quality of models and modeling languages in the information systems field. The evaluation is based on both practical experiences and evaluations of UML and Rational Rose made by others. Based on the evaluation, we conclude that, although being an improvement over its predecessors, UML still has many limitations and deficiencies. Also Rational Rose only partly supports the development of information system models of high quality, and provides too limited support for using different modeling techniques in concert within a larger methodological framework.

The Whole-Part Relationship in the Unified Modeling Language

This study of the semantics of the Whole-Part relationship in OO modeling is based upon, extends and, specifically, formalizes earlier analyses of the semantics of UML’s Aggregation and Composition (white and black diamonds, also called shared aggregation and composite aggregation). Although UML is nowadays regarded as a standard and is widely used as an OO modeling language, the way the Whole-Part relationship is formalized is unsatisfactory. Here, we provide a rigorous specification of various forms of the Whole-Part relationship using OCL (Object Constraint Language). The first part of the specification is based on the differentiation between primary characteristics (applicable to all Whole-Part relationships) assigned to a new Whole-Part metatype in the UML and secondary features, which are possessed by subtypes of this metatype and permit the representation of several “flavors” of the Whole-Part relationship. This UML-compliant style of specification, based on the use of OCL as well as metamodeling, allows us to directly incorporate our results into the UML metamodel, in particular revising UML’s definition of Composition.

Supplementing UML with Concepts from ORM

The Unified Modeling Language (UML) is useful for designing object-oriented code, but is less suitable for conceptual data analysis. Its process-centric use-cases provide an inadequate basis for specifying data-centric class diagrams, and the UML graphical language suffers from incompleteness, inconsistency and unnecessary complexity. For example, multiplicity constraints can lead to unexpected problems when extended to n-ary associations, the constraint primitives are not optimized for orthogonality or expressibility, and the graphical language does not lend itself readily to verbalization and multiple instantiation for validating models with domain experts. This chapter examines some of these defects, and shows how to compensate for them by augmenting UML with concepts and techniques from the Object Role Modeling (ORM) approach. It highlights the potential of “data use cases” for seeding the data model, using verbalization of facts and rules with positive and negative examples to facilitate validation of business rules. The following approaches are suggested as possible ways to exploit the benefits of fact-orientation: use ORM for the conceptual analysis then map the ORM model to UML; use UML supplemented by informal population diagrams and user-defined constraints; enhance the UML metamodel to better support business rules.

Rational Unified Process and Unified Modeling Language - A GOMS Analysis

GOMS is a model that analyzes knowledge of how to do a task in terms of Goals, Operators, Methods and Selection rules. GOMS is one of the most popular theoretical models in the field of human-computer interaction. Since its introduction, the GOMS model has been extended, enhanced and applied to areas outside human-computer interaction. The goal of this chapter is to discuss the use of the GOMS model for the design and evaluation of modeling techniques. In this chapter, we introduce the GOMS concepts, discuss the applicability of GOMS for modeling and describe how GOMS can be used to analyze Rational Unified Process and Unified Modeling Language.

Systematic Design of Web Applications with UML

We propose a systematic design method for Web applications which takes into account the navigation space and the presentational aspects of the application. The method is based on a UML profile for the Web domain. Starting with a use case analysis and a conceptual model of the application, we first provide guidelines for modeling the navigation space. From the navigation space model we can derive, in a subsequent step, a navigational structure model which shows how to navigate through the navigation space using access elements like indexes, guided tours, queries and menus. Finally, a static and dynamic presentation model is constructed. The different models of the design process are represented by using a Web extension of UML. The strength of the presented methodology is that most steps can be performed in a semiautomatic way, thus providing the basis for a systematic mechanism for Web design.

UML Modeling Support for Early Reuse Decisions in Component-Based Development

Component-based development is the software industry’s latest answer to some long-standing problems in software development. Its aim is to make actual reuse of existing software units (components) a widespread reality. We argue that significant reuse decisions can occur in the early stages of system development. Using a theory of the modeling process, we show that choice of abstractions and notations is critical. We investigate the kinds of models that would support early reuse decisions. We show that easily composable, business-oriented abstractions about software behaviour and a functional mental model are necessary. Evaluation of the UML in these terms emphasizes its bias towards structural mental models and abstractions derived from the software domain.