- 學位論文
應用FFP-Growth進行模糊關聯規則之探勘
Application of FFP-Growth to Data Mining by Fuzzy Association Rules
若您是本文的作者，可授權文章由華藝線上圖書館中協助推廣。
摘要
資料庫為現今普遍儲存資料的工具，對於如此龐大與激增的資料而言 ，則可運用資料探勘之技術來進行分析、統整，以探勘出有用的資訊，進而形成知識。在資料探勘技術中，是以關聯規則最常見，其主要是從資料庫中找尋項目之間的關聯性。然而在傳統的關聯規則演算法中，需要不斷地重覆檢查資料庫的候選項目集以及掃瞄資料庫，進而無法提升探勘的效率。 本研究以FP-Growth為基礎，並以模糊分割法相結合，提出一項模糊頻率樣式成長（Fuzzy Frequent Pattern Growth, FFP-Growth）演算法。藉由FP-Growth的壓縮資料與掃瞄資料庫兩次之特性，並以模糊分割法來決定於每一項目之模糊集合，來進行關聯規則之探勘。其特色在於交易資料庫經過更新後，完全不須重新掃瞄原始資料庫就可以產生所有的高頻項目組，則可提升探勘之效率。 在論文第三部份，本研究從數量化的交易資料中探勘模糊關聯規則；另外，近期網際網路的應用成長快速，對於資訊系統上的應用有著重大之影響，故在論文的第四部份，是以從網站伺服器日誌資料中去探勘出有用的網頁瀏覽樣式。而所探勘出來的知識，以期能助於決策者制定行銷策略或經營規則，進而開創出新的商業契機。
並列摘要
Database is the common tool for data archive nowadays. It is valuable to obtain useful data and create new knowledge using data mining technologies to analyze and integrate the tremendous and ever-increasing data. The relational rule method which finds relations in database is the most widespread among the data mining technologies. However, it takes repetitive checking and scanning of the database using traditional relational rule algorithm, and consequently limited the data mining efficiency. This research adopted the FP-Growth as basis and integrated fuzzy partition to propose the Fuzzy Frequent Pattern Growth (FFP-Growth) algorithm that integrated the FP-Growth double database compression and scanning and the fuzzy partition rule to determine the fuzzy group for each item. The feature of this algorithm is that the generate frequent patterns can be created after updating trade database, without re-scanning the original database and therefore improve the data mining efficiency. The fuzzy association rules were investigated from the quantitative trade data in the third part of the research. In addition, due to the fast increase of internet applications, the useful web browse style was explored from web server browsing records. The explored knowledge can be used in marketing and management decision making, and create new business chances.
參考文獻
[36] 胡宜中，使用模糊分割自概念層級架構中找出關聯規則，資訊管理學報，13(3)，2006, pp.63-80。
[37] 黃仁鵬等，高效率之遞增式資料探勘演算法－ICT，電子商務學報，8(3)，2006, pp.393-414。
[3] A. Ragel, B.Cremilleux, “MVC—a preprocessing method to deal with missing values, ” Data & Knowledge Engineering Volume: 18, Issue: 3 , 1996，pp. 189-223.
[6] Bezdek, J. C.(1981), Pattern Recognition with Fuzzy Objective Function Algorithms, Plenum, NY, 1981.
[10] Fayyad, U., Piatetsky—Shapiro, G. and Smyth, P., “The KDD process for Extracting Useful Knowledge from Volumes of Data, ” Communications of The ACM, Volume 39, Number11,1996 pp.27-34.