Aspect mining using relative reduced concept lattice

Crosscutting concerns cannot be well modularized in object-oriented software. The implementation of a crosscutting concern is typically scattered over many locations and tangled with the implementation of other concerns. The presence of crosscutting concerns is one of the major problems in software understanding and evolution. Aspect-oriented programming offers mechanisms to factor them out into a modular unit, called an aspect. Aspect mining tries to identify crosscutting concerns in legacy systems and thus supports the adaptation to an aspect-oriented design. This paper presents an automatic static aspect mining approach that relies on the relative reduced concept lattice. It uses method call tree to describe the relationship between class methods. The method call trees are then subjected to concept analysis. In the resulting relative reduced concept lattice, candidate aspects are detected. An experimental evaluation shows that the approach has a higher automation degree and faster mining rate.