none

• Main Authors: Neng-Hsin Su, 蘇能歆
• Other Authors: Han-Hsiang Wang
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/13576279292824644231

碩士 === 國立中央大學 === 營建管理研究所 === 104 === Construction quality defects significantly affect a construction project by presenting poor appearance, delaying the project and even threatening workers or users’ lives. Promoting quality, therefore, becomes a top priority for a construction project. Because contractors’ quality is the most critical factor contributing to projects’ quality, projects’ quality can be promoted by improving contractors’ quality and benchmarking is a quality management technique for contractors’ quality improvement. Benchmarking has been widely used in different industries to improve company-wide and department-wide performance, and has also been proved in academic research to be effective in promoting quality. Selecting best practices (i.e., target companies to be learned) for a learning company (i.e., a company to learn from the best practices) is an important step in a benchmarking process; however, current methods for selecting best practices in literature have certain insufficiencies. First, these methods are mostly qualitative and hence, the results of best practice selection are subjective and vulnerable to human misjudgment. Second, even though some methods are quantitative but their considerations for evaluating best practices are not suitable for the nature of benchmarking. Last, current methods did not consider the characteristics of construction industry. There, this research proposes a similarity-based approach to benchmarking best practice selection. The proposed approach considers 22 selection factors that are determined through literature review, questionnaire survey and expert interviews and represent the characteristics of construction companies. Two similarity measurement techniques, Euclidean distance and cosine similarity, are adopted in the proposed approach to measure the similarity between a learning company and any of the best practice candidates; the best practice candidates are then prioritized according to the similarity values. Two case studies are conducted to validate the proposed approach. The differences between the measurement results by using traditional subjective decision and the proposed approach in the case studies are analyzed, and possible benchmarking learning strategies are also suggested according to the analysis results and the original data. The research results show that the proposed approach can benefit a benchmarking process by facilitating selecting benchmarking best practices, and further improve the efficiency of this important step.