Earticle

Software development can be improved from many perspectives by combining aspect orientation and model-driven engineering techniques. At a higher level, they can be integrated in two different ways: (1) by handling specifics of aspect orientation at modeling level and later generating object-oriented code, or (2) by transforming an aspect model directly into aspect-oriented code. The latter approach has been shown to have advantages over the former. Consequently, different aspect-oriented code generation approaches have appeared in literature. This paper comparatively evaluates some of those existing approaches, which are well-published and can be used in integration with model-driven engineering process. The results of this study indicate that in order to provide this integration, existing aspect code generation need to be improved from various perspectives. Moreover, these results also provide some insight into the prerequisites for a valuable integration of these approaches into a model-driven engineering process.

The integration of aspect oriented modeling approaches with model-driven engineering process achieved through their direct transformation to aspect-oriented code is expected to enhance the software development from many perspectives. This study aims to assess the existing UML-based aspect- oriented modeling techniques from the perspective of their suitability with regards to integration into model-driven engineering process through aspect-oriented code generation. For this purpose, we selected two mature aspect-oriented modeling approaches, Reusable Aspect Models and Theme/UML, and proceeded to evaluate them in a detailed way from the specific perspectives of design and its mapping to the implementation code. The in-depth comparison reveals some points equally shared by both approaches, and identifies some areas where one has advantage over the other. The study concludes that the Reusable Aspect Models approach may be seen as a preferred approach to handling the task of integration using aspect-oriented code generation.

Along with the boom of Mobile services and the thriving Model Driven Architecture (MDA), we must consider the growing significance and utility of modeling in the development of software and solutions. The main advantages of MDA are the ability to transform one PIM into several PSMs, one for each platform or technology in which the final system will be deployed, and the automatic code generation that implements the system for those platforms from the corresponding PSMs. Service-oriented architectures (SOA) are also touted as the key to business agility, especially when combined with a model-driven approach. Model-Driven Architecture (MDA) is a well-developed concept that fits well with SOA, but until now it has been a specialized technique that is beyond practical application scope of most enterprises. Model driven and Domain engineering is the foundation for emerging “product line” software development approaches and affects the maintainability, understandability, usability, and reusability characteristics of family of similar systems. However, the existing domain engineering methods do not elicit information necessary for the Component-based Software Development process, such as, selecting and configuring appropriate components. In this paper, we suggest a method that systematically defines, analyzes and designs a crop monitoring frameworks domain to enhance reusability effectively in Model Driven Engineering and Component-based Software Development. For this, we extract information objectively that can be reused in a crop monitoring domain from the requirement analysis phase. We sustain and refine crop monitoring information, and match them to artifacts of each phase in domain engineering. Through this method, reusable domain components and malleable domain architecture can be produced. In addition, we show the practical applicability and features of our approach via a new clouding and IOT technology.

Research and practice for Model Driven Engineering (MDE) have significantly progressed over the last decade for dealing with the increase of complexity within systems during their development and maintenance processes by raising the level of abstraction using models as information storage. New significant approaches, mainly Model Driven Architecture (MDA) defined at the OMG (Object Management Group), “Software Factories” proposed by Microsoft and the Eclipse Modeling Framework (EMF) from IBM, are born and have been experimented. As models grow in use for developing systems, transformation between models grow in importance. University and industry are seeking for effective and efficient ways to treat transformation as first-class assets in MDE. In order to produce new and more powerful transformations, we argue that the semi-automatic generation of transformation rules is an important challenge in future MDE development to make it easier, faster, and cost-reduced process. In this paper we propose an extended architecture that aims to semi-automate the process of transformation in the context of MDA. This architecture introduces mapping and matching as first class entities in the transformation process, represented by models and metamodels. We will introduce and discuss briefly two main operations “adaptation” and “derivation” which we consider as core techniques for a semi-automatic transformation process in MDA, along with the first two main techniques of matching and mapping. We will introduce a still theoretical idea already adopted in schema matching field of combining many techniques of matching taken from the literature in order to improve the quality of the obtained mappings. This combination will raise the semi-automation level and thereby significantly contribute to results of higher quality. Finally, an illustrative example is presented and related works are discussed.

Model-Driven Engineering (MDE) tries to reduce the effort spent on software development by generating codes from models. People concentrate their minds on the transformation between models and models, or between models and codes. People also concentrate on checking consistency between different models such as consistency between a class diagram and a sequence diagram and consistency between a sequence diagram and a state machine diagram. Checking relationship consistency and class redundancy in a class model is still important but ignored in recent years. This paper concentrates on relationship problems between classes in a class diagram and proposes methods of checking various relationship problems. We address the redundancy of a class’s operations and attributes. We identify a large range of the problems for class diagram. Our research is based on the relationship abstraction rules.

모델 기반 개발방법은 개발자가 추상화된 모델만을 설계하는 것만으로도 소프트웨어를 개발할 수 있도록 하는 방법이다. 현재까지의 모델 기반 개발방법론은 모델에서 코드를 변환하는 것은 다루고 있지만, 반대로 코드에서 모델로의 변환은 고려하고 있지 않다. 본 논문에서는 모델이 아닌 기존에 작성된 C 언어 코드를 모델로 변환하는 역공학 기법을 제안한다. 이러한 역공학 기법을 사용하면, 새로운 모델을 작성할 때 기존의 코드로부터 모델을 얻어내 적용할 수 있다. 또한, 모델을 작성하고 작성된 모델을 통해 생성된 최종코드를 수정하였을 경우 역공학을 통해 모델과 수정한 코드를 일관성 있게 유지할 수 있다. 이를 지원하기 위해 C 언어를 UML 로 변환하는 방법 및 변환된 모델의 효율적인 구성을 위한 모델 재구성 방법을 제안한다.

모델 기반 개발방법은 개발자가 추상화된 모델만을 설계하는 것만으로도 소프트웨어를 개발할 수 있도록 하는 방법이다. 현재까지의 모델 기반 개발방법론은 모델에서 코드를 변환하는 것은 다루고 있지만, 반대로 코드에서 모델로의 변환은 고려하고 있지 않다. 본 논문에서는 모델이 아닌 기존에 작성된 C 언어 코드를 모델로 변환하는 역공학 기법을 제안한다. 이러한 역공학 기법을 사용하면, 새로운 모델을 작성할 때 기존의 코드로부터 모델을 얻어내 적용할 수 있다. 또한, 모델을 작성하고 작성된 모델을 통해 생성된 최종코드를 수정하였을 경우 역공학을 통해 모델과 수정한 코드를 일관성 있게 유지할 수 있다. 이를 지원하기 위해 C 언어를 UML 로 변환하는 방법 및 변환된 모델의 효율적인 구성을 위한 모델 재구성 방법을 제안한다.