結合自助式商業智慧技術之敏捷資料分析方法－以公部門為例

大數據時代的來臨，導致商業智慧系統再次為大眾所重視，而增強圖形化互動分析的自助式商業智慧系統補足了原有商業智慧系統的不足之處，讓使用者更輕易地自行進行資料分析。然而一般使用者如何和資訊人員合作因應不同的分析目的快速敏捷擷取不同來源的資料，進行整合分析與設計？不同於之前的商業智慧系統開發方法，本研究將敏捷分析做為開發方法建構之參考框架，以快速滿足使用者端需求，並且簡化傳統建置商業智慧系統時需要優先規劃資料倉儲的困擾。本文所提出的敏捷自助式商業智慧系統開發方法包含以下特性：（1）可適用於快速導入自助式商業智慧系統；（2）應用資料庫反向工程方法縮短分析時程；（3）可有效減輕資訊人員負擔；（4）利用故事情境法釐清不同使用者角色定義與需求；（5）使用衝刺計畫與待辦清單有效控管與追蹤開發進度。

關鍵字

自助式商業智慧系統；敏捷分析；視覺化分析

並列摘要

In the era of big data, leading to widespread use of business intelligence systems. Self-service business intelligence system that is enhanced by the function of visualization to overcome the business intelligence system deficiencies, allows users to easily analysis data by themselves. For the challenges of the rapid changes in business environments and system requirements, how do users collaborate with IT staff to develop the self-service business intelligence system to quickly adapt with diverse data from different sources? Traditionally building a business intelligence system requires to implement a data warehouse to integrate different sources of data; in this study, we combine agile development methods to quickly satisfy the different requirements of multiple users, and reduce the efforts of building data warehouses. The development method proposed in this study combine with some techniques to get those benefits:（1）fitting for the quick adoption of self-service business intelligence systems;（2）applying database reverse engineering methods to shorten time for system analysis;（3）efficiently reducing the burden of IT staff;（4）using story telling methods to clarify system requirements of different user roles;（5）using sprint and product backlogs to effectively control and trace the schedule of projects.