Mi-Sook Choi

Component-based metrics applying the strength of dependency between classes

Component-Based Development (CBD) is an emerging discipline for promoting practical reuse of software. In CBD, by building new software with independently developed components, we can gain the benefits promised by the software reuse such as quality improvement and rapid development. Accordingly, to improve quality of components, we propose the component-based metrics applying the strength of dependency between classes to measure precisely. In addition, we prove the theoretical soundness of the proposed metrics by the axioms of Briand et al. and suggest the accuracy and practicality of the proposed metrics through a comparison with the conventional metrics in component development phase.

A Dynamic Coupling for Reusable and Efficient Software System

Recently, dynamic behaviors in embedded system for ubiquitous environment and service-oriented system have needed a dynamic coupling metric to evaluate the quality of the software system more accurately. The embedded systems are designed by a component-based system and the component-based system is designed by one object-oriented. Therefore, coupling between the classes in the object-oriented model are brought into relief more clearly. But, in spite of a great deal of study and effort, most coupling metrics between classes suggest metrics that measure the coupling based on a static dependency relationship. In this paper, we propose a dynamic coupling metric in order to measure the coupling accurately between classes allowing the dynamic property in the object level. In addition, we prove the theoretical soundness of the proposed metric by the axioms of Briand et at. and suggest the accuracy of the proposed metric through a comparison with conventional metrics.

A component identification technique from object-oriented model

Todays software system environment requires rapid development and high productivity. In order to satisfy these requirements, researches have been working on the development of software reuse technology. Therefore, it is general to develop component-based software due to the advance of reuse technology nowadays. One of main issues raised in component-based development (CBD) is how to identify reusable and independent components. Existing methodologies have dealt with the problem based on only developers heuristics, so that it is difficult to identify the components by common developers. Therefore, in this paper, we propose a new technique to identify the business components based on system component. The proposed technique applies the characteristics and degree of dependency between classes in object-oriented model. We also present a case study and experimental results to prove the practical use of our technique. We examined various examples to get the objective foundations.

A component cohesion metric applying the properties of linear increment by dynamic dependency relationships between classes

In Component-Based Development(CBD), by building new software with independently developed components, we can gain the benefits promised by the software reuse such as quality improvement and rapid development. Hence, methods to design the components in component-based system are very important and the component architecture should consist of highly cohesive components. Accordingly, to design highly cohesive components in component development phase, we propose new types of dynamic dependency relationship between classes and a component-based cohesion metric applying the properties of linear increment by dynamic dependency relationships between classes. In addition, we prove the theoretical soundness of the proposed metric by the axioms of briand et al. and suggest the accuracy and practicality of the proposed metric through a comparison with the conventional metrics.

Component Identification Methods Applying Method Call Types between Classes

Identifying reusable and autonomous components is one of the most important and difficult tasks in developing component-based systems. However, the existing component development methodologies do not provide a clear standard for component identification and depend on the intuition and experience of individual developers. As a result, people with limited expertise cannot easily identify components. This paper proposes methods to identify components based on object-oriented techniques including Use Case diagrams, Class diagrams, and Sequence diagrams. We use two distinct steps in identifying components, that is, system and business component identification. In particular, our proposal considers dynamic dependency characteristics of the method call types and directions. Case study and assessment with the existing identification approaches help to verify the practicality of our proposal.

A coupling metric applying the characteristics of components

A high coupling between components makes it difficult to build the modulation of software and the reuse of components, and to manage the system due to the ripple effect by software change impact. Thus, a coupling metric is required to measure the coupling between components in order to design software effectively. In this paper, we propose an approach to improving the existing component-based coupling metrics by considering the dependency about the structured relationships and the method call types between classes. In addition, we prove the theoretical soundness of the proposed metric by the axioms of briand et al. and suggest the accuracy and practicality of the proposed metric through a comparison with the conventional metrics.

A Model Driven Architecture and Product Line Engineering Technique for Adaptable Contents Service of Ubiquitous Computing : Applying to Vessel U-Safety Monitoring

In ubiquitous environments, the content adaptable services can be dynamically provided to adapt the frequent changes of contexts. These services have common things that the kinds of context factors are limited to ubiquitous environment, though the contexts are flexible. To reuse service architecture can be reasonable for effective adaptable service. In this paper, we design a software architecture with product line techniques for content adaptable applications in ubiquitous environment. Description of product line is to define variation points and their variants, to find out the dependencies between them and to keep the model based architecture, their alternatives.

A Modeling Technique Applying an Object-Oriented Petri Net for Embedded System

Embedded systems that mutually interact with physical space have properties such as time constraint, concurrency constraint, and the control of data/resource flow. The exceptional behaviors of the system that interact with physical space can result in critical damage and cause danger to the operation of an embedded system. Thus, in order to classify and express the exceptional behavior in the design phase, our study focuses on the expansion of Petri Net, which accommodates most of the embedded system¿s properties. The reliance of embedded systems can be increased by specifying the embedded system¿s exceptional behaviors.

Identification of Business Component based on Independence Metric

When constructing a component based system, It is understood that identifying reusable and independent business components is of utmost importance. However, according to conventional component based developing methodologies, most of developers depend on their experience and/or intuition for identification of business components. Furthermore, there are no criteria to evaluate whether the identified business components are more independently defined or not. Therefore, we propose a component identification metrics to apply to component properties In order to complement the difficulties of identifying business components through developers` experience and/or intuition. The metrics defined are the criteria for identifying the business Components and/or for evaluating the Identified components. We propose both a cohesion metric, and a coupling metric, to which component properties are applied, wherein those properties can be understood by high cohesion in, and low coupling between, components. Moreover, we propose an independence metric that can evaluate the degree of independence for a particular component by ratio of the cohesion and coupling of components. The metrics that we propose are applied to case study which demonstrates the identification of more independent business components and the validity of our metrics.