Approaches for Software Metrics Threshold Derivation: A Preliminary Review

Tina Beranic, Marjan Hericko
2017 Software Quality Analysis, Monitoring, Improvement, and Applications  
Knowing the reliable software metrics threshold can contribute to product quality evaluation and, consequently, increase the usefulness of software metrics in practice. How to derive software metrics thresholds is a topic of many researchers, either proposing new approaches, or verifying existing methods on different practical projects. A literature review was conducted to summarise the available knowledge in this area. The aim was to find approaches for software metric threshold derivation and
more » ... extract relevant information, such as programming language for which an approach was proposed and tested, what metrics were derived, and if approaches are automated with supporting tools. Data extracted from 14 primary studies reveal that the majority of studies present new approaches and approaches usually apply statistical analysis for threshold derivation. Primary programming language for which metrics are derived is Java, followed by C, C++ and C#. We also came across some tools used for threshold derivation, but in practise their use was not detected. Metrics derived, purpose of metrics derivation and other information are also summarised in the tables bellow.
dblp:conf/sqamia/BeranicH17 fatcat:muqugleczbasji6ijqltxjgstq