一個針對物聯網的資料分析框架

Abstract

物聯網已逐漸成為資訊產業發展的重要領域，隨著聯網裝置的增加，協助資料的擁有者能有效應用其所帶來的海量資料將是重要的課題。本論文提出一個適用於物聯網的資料分析框架，使用者透過此框架能簡單的彙整、分析物聯網資料。 機器學習在物聯網領域中可被用於情境感知及最佳化物聯網的基礎架構等等，扮演相當重要的角色。本論文所提出的框架提供圖形化引導介面，協助使非專業資料科學家的使用者也能受惠於機器學習技術。 本框架同時針對物聯網資料的特性，簡化建立資料應用程式之開發工作。我們透過兩個實際資料應用範例演示本框架可以有效降低開發難度與所需時間，可被應用於快速建立相關系統原型。

The Internet of Things (IoT) has become one of the most important font for the IT industry. With the ever-increasing amount of connected "Things", helping data owners leverage their massive data becomes a major issue. In this paper, we proposed a new data analysis framework for the Internet of Things, to help users aggregate and analysis IoT data easily. Machine learning plays an import role in IoT. It is used in various ways in IoT, such as provide context-awareness or optimizing IoT infrastructure. Our framework provides a graphical interface to provide guidance, and enables non data scientists using machine learning technologies in their projects. Our framework also utilizes the properties of IoT data, reduce the workload of developing IoT data applications. We demonstrate two real data examples to show the time and efforts reduced by our framework. We believe our framework could be used in making quick prototypes.