Using Formal Concept Analysis to support change analysis

Software needs to be maintained and changed to cope with new requirement, existing faults and change requests as software evolves. One particular issue in software maintenance is how to deal with a change proposal before change implementation? Changes to software often cause unexpected ripple effects. To avoid this and alleviate the risk of performing undesirable changes, some predictive measurement should be conducted and a change scheme of the change proposal should be presented. This research intends to provide a unified framework for change analysis, which includes dependencies extraction, change impact analysis, changeability assessment, etc. We expect that our change analysis framework will contribute directly to the improvement of the accuracy of these predictive measures before change implementation, and thus provide more accurate change analysis results for software maintainers, improve quality of software evolution and reduce the software maintenance effort and cost.