Aspect Mining Using Concept Analysis

Crosscutting concern is the inherent limitation for object oriented programming, Aspect oriented program- ming is hopeful to become an effective way of solving the problem. A novel method using concept analysis technology to identify candidate Aspects set from object oriented legacy system is provided in this paper. Methods in legacy system are regarded as items and direct method invocation relationships are as features. Candidate Aspects are generated auto- matically after concept clustering. Candidate Aspects reflect the behavior of the crosscutting concerns in the legacy sys- tem, which is useful for the following Aspect reflecting and program understanding. A actual example of bank program proved the usability of concept analysis.