Identifying Developer Positions in AGILE Projects Using BERT Architecture

The paper is a solution based on machine learning for automatically classifying job titles in AGILE software industry projects. Our methodology incorporates fine-tuning techniques to accurately identify job titles from task descriptions. By utilizing the BERT model, fine-tuned on a dataset of job postings in the software industry, we achieve good performance in understanding contextual information and the semantic nuances of job skills and requirements. The results indicate the model's ability to adapt to various industry projects and accurately identify the project contributors title. These findings highlight the potential of transformer-based models in improving the efficiency and effectiveness of project management practices.