Ambient Software Engineering –Moving Towards A Software Engineering Web of Things

点击翻译按钮获取中文摘要

The Internet has evolved over the last two decades from a global (read only) information hub for end-users, to its current form (Web 2.0) with its end-user support for content creation (e.g., blogging, social networking). However, even in its current development stage, Web 2.0 is only a step towards the original vision of what the Internet might become – a Web of Things also known as Web 3.0. Web 3.0 will be built upon several key technologies (e.g., Semantic Web, automated tagging, reasoning and semantic linking), with machines talking to other machines to deal with much of the daily routine work, leaving humans to provide the inspiration and intuition all in a global context. While many domains (e.g., automotive, manufacturing, home automation) have started to make the Web 3.0 an important and integrated part of their future solution space, the software engineering community has yet to embrace this move. The objective of this research program is to introduce Ambient Software Engineering as a cross disciplinary approach, integrating traditional software engineering (evolution) research, with state-of-the art research from domains such as Semantic Web, Natural Language Processing, knowledge modeling and Mining Software Repositories. The approach takes advantage of matured Semantic Web technologies, which support a standardized unified and formal knowledge representation using ontologies and Linked Data of existing heterogeneous knowledge resources. The proposed research will focus on the introduction of information hubs, with facts and knowledge extracted from existing software resources being formally modeled using ontologies and Linked Data representations. In addition the use of automated tagging, machine learning, source code analysis and Semantic Web reasoning will be investigate to further enrich these information hubs with semantic information and semantic traceability links - both being an essential prerequisite for establishing a Software Engineering Web of Things. For the evaluation an initial set of smart workflows and services for the software evolution domain will be introduced that allow for the support of composite applications or task contexts across global heterogeneous knowledge extracted from the software engineering domain. Both, the datasets and smart services being developed, will be publicly available through an online portal, to create an initial version of a Web of Things for the software engineering domain, where information hubs replace traditional information silos found in the software domain. These information hubs (things) will promote knowledge sharing, reuse and integration, by facilitating a crowd-sourcing effort involving the software community at large to share and link their own data (facts and analysis results) and smart services to continue evolving this Web of Things for the software engineering domain.