This article presents a novel approach to the research of data mining database components based on BIM systems in large-scale industrial projects and assembled buildings, particularly in the context of a few similar engineering pro- jects. The study begins at the design and financing stage of the industrial sub- project proposal and extends to the preliminary design stage. It addresses the challenge of multi-party cost requirements and proposes a comprehensive process of BIM-based data requirements, extraction, sorting, mining, preprocessing, and display. The method is validated through a preliminary project test, demonstrating the project's dynamic data collection and output. The study also includes an analysis of typical industrial plant projects from four perspectives: sub-item characteristics, sub-item project prices, cost components, and leading project quantity indicators, leading to developing indicators for subsequent projects.