Akademik Veri Yönetim Sistemi
PAMUKKALE UNIVERSITY JOURNAL OF ENGINEERING SCIENCES-PAMUKKALE UNIVERSITESI MUHENDISLIK BILIMLERI DERGISI, cilt.26, sa.5, ss.993-1014, 2020 (ESCI)
The software industry is increasingly involved in every aspect of life and software projects are being developed to a large extent. This situation causes very important faults and negative results in the developed projects. Therefore, in order to prevent or minimize this situation, software risk management activities must be successfully implemented. In this study, a new web-based risk management process based on artificial intelligence in software projects was designed and developed. The purpose of the study is to estimate the deviations that might occur in the project outputs according to the risk factors using artificial neural networks (ANN), to minimize the harm that may be encountered in the first stages of the software life cycle and thus to provide a preventive approach for users. In order to create the ANN model of the study, a checklist form was created by preliminary discussions with academicians, experts and project managers in the software engineering field. By using this form, the actual project data were collected from 774 different companies in the software companies located in Teknokent. The generated ANN model has forty-five entrances, a single hidden layer (with fifteen neurons) and five outlets (with 45-15-5); the education R rate is 0.9978; the test R ratio is 0.9935 and the error rate is 0.001. The model is integrated into the application developed by creating the.dll library. The developed application, real project data from different areas were obtained and after obtaining the opinions of experts and academicians (10 people), 4 different scenarios were tested and results were obtained. The results clearly demonstrate that the performance of the application is high and that the use of ANN in such applications provides positive contributions to the project's success. In addition, it has been found that there is a need for applications that provide an artificial intelligence-based risk management process for the software industry.