查詢結果分析
來源資料
頁籤選單縮合
題 名 | 環境與災防海量資料應用系統之架構、範例與挑戰 |
---|---|
作 者 | 蔡惠峰; 李隆正; 林聖峰; 張文鎰; 林錫慶; | 書刊名 | 國土資訊系統通訊 |
卷 期 | 88 2013.12[民102.12] |
頁 次 | 頁32-45 |
專 輯 | Big Data 與 GIS |
分類號 | 312.13 |
關鍵詞 | 氣候變遷; 海量資料; 環境變化; 天然災害; |
語 文 | 中文(Chinese) |
中文摘要 | 全球氣候變遷引起之環境變化與天然災害需要進行長期之監測與預測,這是一個 海量資料的生成環境,環境與災防資料具備海量資料之龐大、多元與即時之特性。環 境與災防所屬之地球科學牽涉空間大尺度及時間長延時之監測與預測,因此不僅在監 測產生龐大資料,在高速計算模擬過程中亦產生巨量預測資料,部份資料成果且須進 行即時之展示 (Visualization)、影像特徵分析與資料探勘分析 (Mining),此並非一般傳 統之 Hadoop 平台純為資料管理及分析探勘目的可勝任,因此本文以全球氣候模擬為 例,說明此類大型模擬計算導向之海量資料平台之架構建置之獨特性與推動門檻;另 以橋樑監測為例,陳述長期監測之影像海量資料,所需Hadoop 海量資料平台之架構。 |
英文摘要 | To respond to global climate change, people need to long-term monitor the environmental change and relevant natural disaster events that is inherently the right atmosphere to create the big data. Environmental monitoring and disaster management have the features of big data: large volume, variety, and velocity (data in motion or real-time). Large-scale and long-term monitoring and predicting the phenomena in earth are necessary actions. Not only the sensor network for monitoring creates the big data, but also does the high-performance simulation during the predicting process. During simulation, it may have to display the temporary results in real-time and conduct analytics. This type of problem obviously cannot rely on the traditional Hadoop platform solution that purely works for batch data management and analysis. Therefore, this paper takes the example of global climate modeling to illustrate the framework of big data platform for which the big data created by high-performance simulation and requires real-time visualization and image featuring. Another example is the long-term bridge monitoring in river, the Hadoop platform is planned for the big image data analytics. |
本系統中英文摘要資訊取自各篇刊載內容。