Identification of trends in devops posts on the stack overflow platform using tag and topic modeling analysis

In this thesis, a tag analysis and topic modeling analysis are conducted on posts related to the concept of DevOps, which combines software development and operations, on the Stack Overflow (SO) platform. The aim of the study is to identify the most discussed topics related to DevOps, its historical development, and potential future trends. The dataset consists of 30,600 questions and 34,327 answers posted on the SO platform under the "devops" tag between 2008 and 2023. During the analysis process, descriptive statistics and the Latent Dirichlet Allocation (LDA) algorithm were used to examine the data. The study's findings revealed a significant increase in the use of the "azure-devops" tag starting from 2014. Additionally, DevOps-related posts were clustered under a total of 24 main topics, with themes such as "continuous integration," "deployment automation", "application performance management," and "test-driven development" standing out. The results indicate that the DevOps community generally deals with complexities in continuous integration and deployment processes, tool integration, and performance management. To overcome these issues, it is recommended to enhance collaboration and communication, and focus on continuous learning and improvement processes. This study makes significant contributions to the optimization of DevOps practices and the deepening of knowledge in this field.