Augmenting design patterns with design rationale

A precondition for reusability in software development is the existence of reusable resources. The number of reusable resources is limited owing to a de nite lack of systematic methods for producing reusable information. It is this de ciency which prevents e ective software reuse within a collaborative environment. To address this problem, a new methodology for supporting the collaborative development of software applications by a group of software developers from a library of building block cases has been proposed. The new methodology focuses on the use of an explicit software development process to capture and disseminate specialized knowledge that augments the description of the software cases in a library. The importance of preserving and using this knowledge has become apparent with the recent trend of combining the software development process with the software code. There are three main components in the new methodology: design patterns, design rationale and case-based reasoning. Design patterns have been chosen to represent, record and reuse the recurring design structures and associated design experience in object-oriented software development. The Design Recommendation and Intent Model (DRIM) was used in the proposed methodology to capture the speci c implementation of reusable software systems. DRIM provides a method by which design rationale from multiple software designers can be partially generated, stored and later retrieved by a computer system. To address the issues of retrieval, indexing and adaptation of project information, the methodology used case-based reasoning mechanisms for reusing software from previously developed projects taking into account the di erence in contextual information. The present work also details an initial proof-of-concept prototype based on the new methodology. The proposed methodology is expected to provide a strong information base for software understanding by supporting non-obtrusive capture as well as e ective access of design rationale information regarding the software development process. Thesis Supervisor: Feniosky Pe~ na-Mora Title: Assistant Professor Augmenting Design Patterns With Design Rationale by Sanjeev Sureshchandra Vadhavkar Submitted to the Department of Civil and Environmental Engineering on January 16, 1997, in partial ful llment of the requirements for the degree of Master of Science Abstract A precondition for reusability in software development is the existence of reusable resources. The number of reusable resources is limited owing to a de nite lack of systematic methods for producing reusable information. It is this de ciency which prevents e ective software reuse within a collaborative environment. To address this problem, a new methodology for supporting the collaborative development of software applications by a group of software developers from a library of building block cases has been proposed. The new methodology focuses on the use of an explicit software development process to capture and disseminate specialized knowledge that augments the description of the software cases in a library. The importance of preserving and using this knowledge has become apparent with the recent trend of combining the software development process with the software code. There are three main components in the new methodology: design patterns, design rationale and case-based reasoning. Design patterns have been chosen to represent, record and reuse the recurring design structures and associated design experience in object-oriented software development. The Design Recommendation and Intent Model (DRIM) was used in the proposed methodology to capture the speci c implementation of reusable software systems. DRIM provides a method by which design rationale from multiple software designers can be partially generated, stored and later retrieved by a computer system. To address the issues of retrieval, indexing and adaptation of project information, the methodology used case-based reasoning mechanisms for reusing software from previously developed projects taking into account the di erence in contextual information. The present work also details an initial proof-of-concept prototype based on the new methodology. The proposed methodology is expected to provide a strong information base for software understanding by supporting non-obtrusive capture as well as e ective access of design rationale information regarding the software development process.A precondition for reusability in software development is the existence of reusable resources. The number of reusable resources is limited owing to a de nite lack of systematic methods for producing reusable information. It is this de ciency which prevents e ective software reuse within a collaborative environment. To address this problem, a new methodology for supporting the collaborative development of software applications by a group of software developers from a library of building block cases has been proposed. The new methodology focuses on the use of an explicit software development process to capture and disseminate specialized knowledge that augments the description of the software cases in a library. The importance of preserving and using this knowledge has become apparent with the recent trend of combining the software development process with the software code. There are three main components in the new methodology: design patterns, design rationale and case-based reasoning. Design patterns have been chosen to represent, record and reuse the recurring design structures and associated design experience in object-oriented software development. The Design Recommendation and Intent Model (DRIM) was used in the proposed methodology to capture the speci c implementation of reusable software systems. DRIM provides a method by which design rationale from multiple software designers can be partially generated, stored and later retrieved by a computer system. To address the issues of retrieval, indexing and adaptation of project information, the methodology used case-based reasoning mechanisms for reusing software from previously developed projects taking into account the di erence in contextual information. The present work also details an initial proof-of-concept prototype based on the new methodology. The proposed methodology is expected to provide a strong information base for software understanding by supporting non-obtrusive capture as well as e ective access of design rationale information regarding the software development process. Thesis Supervisor: Feniosky Pe~ na-Mora Title: Assistant Professor