文摘
Mining software repositories has become an important activity during software evolution,as the extracted data is used to support a variety of software maintenance tasks. The key information extracted from these repositories gives a picture of the changes on the software system. To have a complete picture,tailored to the needs of the developer,the extracted data needs to be filtered,aggregated,and presented to the users. In the thesis we propose a new visualization tool for such data,named CVSViewer3D cv3D). The new visualization allows users to define multiple views and perspectives of the change history data,extracted from CVS repositories,and each view helps answer a set of questions relevant to specific maintenance tasks. Data can be viewed at different granularity e.g.,file,line of text,method,class) and comprehensive views can be defined,which display to the user multiple data types at the same time. Complex questions and tasks can be solved with the aid of such views.