A Strategy to Support Replications of Controlled Experiments in Software Engineering
A Strategy to Support Replications of Controlled Experiments in Software Engineering
复制标题
支持软件工程中受控实验复制的策略
DOI：
--
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
Amadeu Anderlin Neto
中科院分区：
文献类型：
--
作者：
Amadeu Anderlin Neto
复制对于建立经验科学知识至关重要。在软件工程背景下报告的实验重复在其实验元素上呈现可变性，例如，变量和材料进一步了解这些变异性，可以帮助规划缺乏策略来支持的实验变异性和共性的表征。此外，在实验要素的有效再利用和可追溯性方面也存在差距。这些问题很可能妨碍对复制的理解和规划。为了克服这些差距，我们打算创建一个概念模型和一个工具，以支持复制规划。为了开发这些解决方案，我们将使用实验和软件产品线的概念。我们的想法是建立一个核心结构，它允许配置的实验元素的基础上与以前的复制和所需的变异性，以适应特定的复制目的的共性。在本文中，我们描述了相关的工作，我们的研究方法，以及目前的研究进展和预期的未来贡献。
Replication is essential to build knowledge in empirical science. Experiment replications reported in the software engineering context present variabilities on their experiment elements, e.g., variables, materials. Further understanding these variabilities could help planning lack of strategy to support the representation of experiment variabilities and commonalities. In addition, there is also a gap related to effective reuse and traceability of experiment elements. These problems are likely to hamper the replication understanding and planning. In order to overcome these gaps, we intend to create a conceptual model and a tool to support replication planning. To develop these solutions, we will use concepts of experimentation and software product lines. Our idea is to build a core structure which allows the configuration of experiment elements based commonalities with previous replications and desired variabilities to fit the specific replication purposes. In this paper we describe related work, our research methodology, as well as the current research progress and expected future contributions.