Software Maintainability Datasets collection across Android and Apache Kafka Versions

High-quality software production is contingent upon software maintainability. Various methods have been suggested to assess software maintainability in a quantitative manner. Software metrics, such as the no. of LOC and McCabe’s Cyclomatic Complexity(CC), Chidamber & Kemerer (C & K) metrics and Lorenz & Kidd( L & K) metrics are significantly used in such approaches. In this study, Five Software Maintainability datasets, namely Android Calendar(small size), Android Telephony(medium size), Android Services(small size), Android Net(medium size) and Apache Kafka(large size) are calculated which are based on Android and Apache based Open Source software written in Java that contain information, i.e., various OO metrics as Independent variables for every class present in the source code and change(Alter) as Dependent variable to indicate whether the class is altered or not. Various Software Metrics Collection Tools for Java projects are analyzed and found out that Understand: The Software Developer’s Multi-Tool(UnderStand for Java) by SciTools calculate OO Metrics that focuses on every aspect of software development, i.e, coupling, cohesion, size and inheritance, as well as LOC and CC which other software’s lack in generating certain metrics related to Software development. UnderStand for Java Tool was employed to extract OO metrics from versions of Android Calendar 9.0.0, Telephony 9.0.0, Services 13.0.0, Net 13.0.0, and Apache Kafka 3.6.0. WinMerge was used to calculate the binary Alter variable (0 or 1) by comparing these versions with their subsequent releases: Android Calendar 10.0.0, Telephony 10.0.0, Services 14.0.0, Net 14.0.0, and Apache Kafka 3.7.0. Point Biserial correlation is used to identify the relationship between OO Metrics and Alter variable thus in turn are indicators of Software Maintainability. Data Processing techniques are applied on these collected datasets to remove bias, maintain data integrity and enhancing SMP. The final datasets, encompassing the evolutionary changes between software versions, are Android Calendar 9.0.0 r1 to 10.0.0 r1, Android Telephony 9.0.0 r1 to 10.0.0 r1, Android Services 13.0.0 r1 to 14.0.0 r1, Android Net 13.0.0 r1 to 14.0.0 r1, and Apache Kafka 3.6.0 r1 to 3.7.0 r1 which are collected for enhancing Software Maintainability Prediction(SMP). This study aims to provide a structured methodology for assessing software maintainability through the correlation of OO metrics and version-based alterations, facilitating improved SMP models.