thesis - Faculty of Information and Communication Technologies ...
thesis - Faculty of Information and Communication Technologies ...
thesis - Faculty of Information and Communication Technologies ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Chapter 1. Introduction<br />
compute at the level <strong>of</strong> a version allow us to identify classes that have<br />
been added, removed, modified <strong>and</strong> deleted between versions. Class<br />
level change measures allow us to detect the magnitude <strong>and</strong> frequency<br />
<strong>of</strong> change an individual class has undergone over its lifetime within the<br />
s<strong>of</strong>tware system. We use the information collected to derive a set <strong>of</strong><br />
common statistical properties, <strong>and</strong> identify if certain properties within<br />
a class cause them to be more change-prone.<br />
1.3 Main Research Outcomes<br />
In this <strong>thesis</strong> we address the problem <strong>of</strong> identifying, in successful s<strong>of</strong>tware<br />
systems, where <strong>and</strong> how maintenance effort tends to be devoted.<br />
We show that maintenance effort is, in general, spent on addition <strong>of</strong> new<br />
classes with a preference to base new code on top <strong>of</strong> a small set <strong>of</strong> class<br />
that provide key services. Interestingly, these choices make the heavily<br />
used classes change-prone as they are modified to meet the needs <strong>of</strong> the<br />
new clients.<br />
This <strong>thesis</strong> makes a number <strong>of</strong> significant contributions to the s<strong>of</strong>tware<br />
evolution body <strong>of</strong> knowledge:<br />
Firstly, we investigated the validity <strong>of</strong> Lehman’s Laws <strong>of</strong> s<strong>of</strong>tware evolution<br />
related to growth <strong>and</strong> complexity within our data set, <strong>and</strong> found<br />
consistent support for the applicability <strong>of</strong> the following laws: First law<br />
Continuing Change, third law Self Regulation, fifth law Conservation <strong>of</strong><br />
Familiarity, <strong>and</strong> the sixth law Continuing Growth. However, our analysis<br />
was not able to provide sufficient evidence to show support for the<br />
other laws.<br />
Secondly, we investigated how s<strong>of</strong>tware metric data distributions (as<br />
captured by a probability density function) change over time. We confirm<br />
that s<strong>of</strong>tware metric data exhibits highly skewed distributions, <strong>and</strong><br />
show that the use <strong>of</strong> first order statistical summary measures (such as<br />
mean <strong>and</strong> st<strong>and</strong>ard deviation) is ineffective when working with such<br />
data. We show that by using the Gini coefficient [91], a high-order<br />
statistical measure widely used in the field <strong>of</strong> economics, we can inter-<br />
6