An adequate complexity metrics set for large-scale
OO software systems is still a challenge. In traditional
OO software engineering methodologies, OO metrics
such as CK and MOOD have been recognized in
practical software development. In order to measure a
system’s complexity at different levels, we propose a
hierarchical complexity metrics set that integrates with
these metrics and parameters of complex networks.
Then, we prove the feasibility of our set through
analyzing the data from SCRR platform.
Although the complexity can be measured in terms
of three categories of metrics, the broad validity of our
set demands further demonstration. So, the future work
includes: 1) investigating the relationships among
different metrics through the regression analysis of
open source OO systems; and 2) design patterns are and change, so the metrics for their complexity would
provide insight into the design quality.