Ambient software engineering

点击翻译按钮获取中文摘要

Over the last decades, software engineering processes have evolved from their industrial roots to include new social and collaborative aspects such as pair-programming and shared virtual workspaces using new technologies such as Web 2.0 (e.g., communities, blogs, services, and wikis). Preservation of relevant software artifacts (e.g. work products and documentation) and their formal and informal records is of critical importance. These artifacts reflect a significant investment made by organization and software communities, capturing (domain) knowledge and expertise that is essential to ensure the evolvability and quality of software systems. At the same time, mining techniques applied on software repositories have matured and are now capable of providing a wealth of information collected from existing software projects. However, due to constant changes made to project teams and software systems, existing project knowledge and expertise is often no longer complete, consistent, or traceable to its origins, resulting in individual repositories that can be considered information silos. In the proposed research, we focus on the need to capture, integrate, and analyze these large and often highly distributed knowledge silos. The primary goal of this research is to create an ambient software engineering environment that facilitates a common knowledge representation to allow a semantic analysis across the boundaries of individual knowledge resources in order to support the re-establishing of traceability links among these existing knowledge resources. The proposed research will address several important issues: (1) the development of (meta-) models in order to support integration, traceability, and reuse among already existing knowledge resources; (2) the development of an ambient software engineering environment that will enable the management, exploration, and reasoning across knowledge and resources boundaries to support software evolution activities through contextual guidance, semantic analysis, quality assessments, and process improvement; and (3) the evaluation of the impact and efficacy of the approach through, metrics, user feedback and case studies.