Ginger2: an environment for computer-aided empirical softwareengineering

摘要

Empirical software engineering can be viewed as a series ofnactions to obtain knowledge and a better understanding about somenaspects of software development, given a set of problem statements innthe form of issues, questions or hypotheses. Experience has made usnaware of the criticality of integrating the various types of data thatnare collected and analyzed as well as the criticality of integrating thenvarious types of activities that take place, such as experiment designnand the experiment itself. This has led us to develop a Computer-AidednEmpirical Software Engineering (CAESE) framework to support thenempirical software engineering lifecycle. The paper first presents thenCAESE framework that consists of three elements: (1) a process model fornthe “lifecycle” of empirical software engineering studies,nincluding needs analysis, experiment design, actual experimentation, andnanalyzing and packaging results; (2) a model that helps empiricalnsoftware engineers decide how to look at the “world” to benstudied in a coherent manner; (3) an architecture, based on which CAESEnenvironments can be built, consisting of tool sets for each phase of thenprocess model, a process management mechanism, and the two types ofnintegration mechanism that are vital for handling multiple types ofndata: data integration and control integration. Next, the paperndescribes the Ginger2 environment as an instantiation of our framework.nIt concludes with reports on case studies using Ginger2, which dealtnwith a variety of empirical data types including mouse and keystrokes,neye traces, 3D movement, skin resistance level, and videotapedndata