文档介绍：Analyzing Empirical Data in Software Engineering
Li Jiang
Armin Eberlein
Aneesh Krishna
School puter Science
The University of Adelaide, SA, 5000, Australia
Computer Engineering Department
American University of Sharjah, UAE
Curtin University of Technology Perth, WA 6102, Australia
Abstract: Getting meaningful information from empirical data is a challenging task in software engineering (SE) research. It requires an in-depth analysis of the problem data and structure to select the most suitable data analysis methods as well as an evaluation of the validity of the analysis result. This paper reports experiences with three data analysis methods that were used to analyze a set of empirical data. One of the major findings is that although each method has its own value, none of them is sufficient to address all challenges on its own. The research reveals that it is only possible to get meaningful analysis results if several data analysis methods bined.
Keywords: Requirements Engineering, Software Engineering, Requirements Engineering Techniques, Data Analysis Methods, Clustering.
Introduction
The development of large and medium-sized software systems usually plex processes that make use of several development techniques. Since the term “software engineering (SE)” was first coined in 1968 at the first SE conference, numerous SE techniques have been proposed. However, e