ABSTRACT
Risk management in software engineering projects describes an integrated design to prevent project failure with methods, processes, and artifacts that continually identify, analyze, control, and monitor risks. For example, changes in people's lifestyles during the COVID-19 pandemic pose unexpected risks to the information technology industry. Agile is known as a methodology that is responsive and adapts quickly to change. Scrum is the most frequently used method based on the 2016 Agile development survey results. Many studies have produced a risk management framework for Scrum in recent years. However, repeating the risk analysis process and selecting a response to risk becomes a burden for stakeholders, so a framework is needed that can become a support system to help make decisions. This paper used a comparative study of risk management framework literature and literature that utilizes risk management tools and a case study of risk classification using 34k GitHub Issues for data mining. This study proposed a new framework that integrates datasets and machine learning into a risk management framework. The novelty in this paper is that the risk priority scheme is carried out using Long Short-Term Memory (LSTM) and Multinomial Naive Bayes (MNB). Further analysis can be carried out to test the overall effectiveness of the framework. © 2022 IEEE.