頁籤選單縮合
題名 | 應用漸增式資料探勘技術於資料倉儲環境之研究=Incremental Data Mining in the Data Warehouse Environment |
---|---|
作者 | 趙景明; 林振群; Chao, Ching-ming; Lin, Chen-chun; |
期刊 | 東吳經濟商學學報 |
出版日期 | 20040300 |
卷期 | 44 2004.03[民93.03] |
頁次 | 頁81-114 |
分類號 | 312.49 |
語文 | chi |
關鍵詞 | 資料倉儲; 漸增式資料探勘; 關聯規則; 決策樹; Data warehouse; Incremental data mining; Association rule; Decision tree; |
中文摘要 | 近年來資料庫領域發展出資料探勘的技術來有效率地從大量的資料中找尋有價值的知識,早期資料探勘的研究以交易資料庫為資料來源,近年來的研究指出資料倉儲是更理想的資料來源,理論上資料倉儲在知識發現,過程 (KDD Process, Knowledge Discovery in Databases Process) 中,可以提供資料探勘一個歷史性、彙整性、整合性與一致性的資料來源,以提昇資料探勘過程的效率與探勘結果的品質。由於資料倉儲環境的特性不同於交易資料庫,使得過去大部份的資料探勘演算法不適合直接套用在資料倉儲環境,因此在本文中,我們探討資料倉儲、資料探勘相關的研究文獻,探究資料倉儲適合做資料探勘的原因及資料倉儲環境的特性,再利用這些特性,提出漸增式資料探勘的方法,並善用資料倉儲環境的工具,以提升在資料倉儲環境下進行資料探勘的效能。 |
英文摘要 | In the past few years, the database research community has developed data mining technology to extract valuable knowledge from massive datasets efficiently. Earlier works on data mining exploit transaction databases as the source of data. Recent literatures indicate that the data warehouse is a superior source of data. In the process of Knowledge Discovery in Databases (KDD), the data warehouse provides a historical, summarized, integrated and consistent source of data for data mining, improving the efficiency and quality of data mining. Because of different characteristics between data warehouses and transaction databases, most algorithms proposed in data mining literatures are not suitable to be implemented in the data warehouse environment directly. In this paper, therefore, we review related works in data mining and data warehousing literatures as well as discuss the need for the integration of data mining and data warehousing technologies and the specific characteristics of data warehouses. By leveraging these characteristics and accommodating the tools of data warehouses, we propose an incremental mining technique to enhance the efficiency of data mining in the data warehouse environment. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。