ABSTRACT
Risk detection and allocation becomes increasingly important to successful project delivery. This is particularly true for mega infrastructure projects, where technical and institutional complexity increases the risk and challenges for collaboration. While early studies developed various methods and tools for risk detection and allocation, industry practice remains experience-based and focuses on opinions and discussions from subject matter experts. This paper will examine the effectiveness of existing methods to identify construction risks and then presents a novel approach to risk detection using case based reasoning and text mining techniques. The method is built on a large project risk database and features semantic inquiry and automatic generation of risk register according to specific project characteristics. I-495/270 managed lanes project from the state of Maryland will be used to demonstrate the process of automatic risk detection.
