本研究以薄膜電晶體液晶顯示器 (Thin-Film Transistor Liquid-Crystal Display, TFT-LCD) 製造業者為研究個案，導入精實六標準差的管理手法─定義、衡量、分析、改善、控制 (DMAIC)，以解決資料倉儲資料的品質問題。研究目的在改善個案公司資料的正確率及完整度，以提升顧客滿意度。DMAIC手法是收集顧客心聲，以定義問題，並找出影響顧客滿意的關鍵品質特性。我們透過資料倉儲的資料處理流程之明細流程圖，來衡量資料倉儲的資料品質現況，以利後續分析。利用特性要因圖找出可能原因，並針對可能原因進行驗證以確認真因。針對問題真因，擬定、實施對策並確認其改善成效。最後，建立監控機制以維持其改善成果。本研究實作後發現DMAIC能有效且快速地提升資料品質。在短短2.5個月，資料倉儲資料的平均正確率由50%提升到81%，其資料的平均完整度由74%提升到90%。In this study, a thin film transistor liquid crystal display (TFT-LCD) manufacturer as the case company, we applied Lean Six Sigma approach—define, measure, analyze, improve, control (DMAIC), to improve the data quality of its data warehouse. The purpose is to improve the accuracy and integrity of the data warehouse, in order to enhance customer satisfaction. DMAIC approach is to collect customer voices, to define the problem, and to identify the critical-to-quality characteristics affecting customer satisfaction. We used the detailed data processing flowchart of data warehousing to identify the data quality problems for further analysis. The cause-effect diagram is then employed to identify possible causes and their verification. With respect to real causes, we developed and implemented action plans to improve data accuracy and integrity. Finally, monitoring mechanisms are established in order to maintain the effectiveness of the lean six sigma approach. We found that DMAIC can effectively and quickly improve data quality of data warehouse since its implementation in March 2012. The average data accuracy of the data warehouse increased from 50% to 81%, and the average data integrity went up from 74% to 90% in 2.5 months.
