Akademik Veri Yönetim Sistemi
Smart Applications with Advanced Machine Learning and Human-Centred Problem Design. ICAIAME 2021, D. Jude Hemanth,Utku Kose,Junzo Watada,Bogdan Patrut, Editör, Springer Nature, Zug, ss.399-407, 2023
One of the most important processes of software development is the presentation of the software in versions. Software versioning allows both developers to provide software management especially dependency control and users to keep track of changes clearly. During the development of the software, the transition to new versions is made according to the type and number of changes made. Managing these transitions and documenting the changes is a costly process. In this context, automating software versioning will make an important contribution to developers. An important step for automatic versioning is the classification of changes in the release notes where the versions are documented. In this study, estimating the type of changes in 800 release notes taken from 4 different projects has been tried with different machine learning techniques. According to the results, it was clearly seen that the SVM method gave the best results in terms of time-performance.