Enhancing RFI Analysis in Construction Projects: A Comparative Study of Text Clustering Methods and Visualization Techniques

Abstract: The Request for Information (RFI) is a vital communication tool in construction projects, aiding teams in addressing queries and navigating challenges. Unstructured RFIs hinder manual analysis for extracting hidden knowledge. Prior research in RFI analysis employed NLP and text clustering and often relied on a single clustering method. This paper performs a comparative analysis using diverse clustering methods (LDA, NMF, and K-means) and visualization techniques to determine the most suitable methods. The study offers project managers and quality engineers an effective tool for extracting hidden knowledge in RFI.