Abstract: The traditional management system for building maintenance work orders is highly likely to ignore the analysis of the textual description part of a work order. Consequently, valuable information is submerged in a large amount of messy data, which makes it difficult to extract repeated and high-frequency work orders quickly and accurately. To solve the above problem, this paper adopted a Chinese word segmentation algorithm based on a keyword library to properly segment the long textual description of repair content in building maintenance work orders. Then, the density detection algorithm based on K-means was employed to introduce the weight of each attribute of the work order and further calculate the weighted Euclidean distance between any two work orders. The density of each work order was obtained, and candidate repeated work order sets were extracted. Finally, the density-based spatial clustering of applications with noise (DBSCAN) algorithm was utilized to determine the final repeated work order set, and the proposed method was applied in an actual project for verification. The results show that the proposed method can accurately and effectively extract repeated work orders from a large amount of data, thereby improving the efficiency of analyzing building maintenance work orders and ultimately ensuring the level of refined logistics management.