在電子商務的市場行銷，決策者會根據不同的策略主題(Strategy Subjects)，例如市場佔有率、最大利潤等等不同的主題來訂定行銷的策略。因此以資料探勘之分群技術進行市場行銷分析時，須考量產品類別、價位及消費者的基本資料等屬性，對於不同行銷策略的權重影響。本研究提出加權式分群法，考量屬性之權重分佈以進行不同策略主題之分析。本文以一電子商務的實例資料集，針對商品的市場佔有率、商品獲利率以及商品推薦為策略主題，進行實驗以驗證加權式分群法之價值

In managing marketing activities, enterprises usually make marketing decisions according to different strategy subjects such as market holding ration or maximum profit. The weightings of features, including product category, unit price and customer data, may vary for different strategy subjects. Accordingly, applying data mining technique such as clustering to market analysis needs to consider the weightings of features. This work proposed a feature-weighting clustering approach to analyze various marketing strategy subjects via considering the weightings of features. Experimental evaluations were conducted to evaluate the effect of the proposed approach on three strategy subjects including market holding ratio, sale revenue and product recommendations.

中文提要 ．．．．．．．．．．．．．．．．．．．．．．．．． i
英文提要 ．．．．．．．．．．．．．．．．．．．．．．．．． ii
誌謝 ．．．．．．．．．．．．．．．．．．．．．．．．． iii
目錄 ．．．．．．．．．．．．．．．．．．．．．．．．． iV
表目錄 ．．．．．．．．．．．．．．．．．．．．．．．．． Vii
圖目錄 ．．．．．．．．．．．．．．．．．．．．．．．．． viii
一、 緒論．．．．．．．．．．．．．．．．．．．．．．．． 1
1.1 研究背景．．．．．．．．．．．．．．．．．．．．．． 1
1.2 研究動機．．．．．．．．．．．．．．．．．．．．．． 3
1.3 問題描述與研究目標．．．．．．．．．．．．．．．．．． 4
1.3.1 異質型屬性．．．．．．．．．．．．．．．．．．．．． 4
1.3.2 資料分佈的趨勢．．．．．．．．．．．．．．．．．．． 4
1.3.3 權重配置．．．．．．．．．．．．．．．．．．．．．． 5
1.3.4 知識的呈現．．．．．．．．．．．．．．．．．．．．．． 5
1.4 研究範圍．．．．．．．．．．．．．．．．．．．．．． 6
1.5 研究步驟．．．．．．．．．．．．．．．．．．．．．． 7
1.6 論文架構．．．．．．．．．．．．．．．．．．．．．． 8
二、 文獻探討．．．．．．．．．．．．．．．．．．．．．． 9
2.1 資料探勘技術．．．．．．．．．．．．．．．．．．．． 9
2.1.1 決策樹與加權分類法．．．．．．．．．．．．．．．．． 9
2.1.2 一般分群技術．．．．．．．．．．．．．．．．．．．． 11
2.1.3 多維度分群．．．．．．．．．．．．．．．．．．．．． 12
2.1.4 人工智慧與類神經網絡．．．．．．．．．．．．．．．． 12
2.2 屬性加權．．．．．．．．．．．．．．．．．．．．．． 13
2.2.1 異質型屬性．．．．．．．．．．．．．．．．．．．．． 13
2.2.2 多變量分析．．．．．．．．．．．．．．．．．．．．． 13
2.3 電子商務與推薦技術．．．．．．．．．．．．．．．．． 13
2.3.1 平衡計分卡(Balanced Score Card, BSC) ．．．．．．．． 13
2.3.2 客戶關係管理(Customer Relationship Management, CRM) 14
2.3.3 推薦技術(Recommendation Systems) ．．．．．．．．． 15
三、 屬性加權的分群技術．．．．．．．．．．．．．．．．． 16
3.1 基本假設．．．．．．．．．．．．．．．．．．．．．． 16
3.2 與分類技術的比較．．．．．．．．．．．．．．．．．． 18
3.2.1 目標導向．．．．．．．．．．．．．．．．．．．．．． 18
3.2.2 資料分佈的劃分．．．．．．．．．．．．．．．．．．． 19
3.2.3 多目標決策．．．．．．．．．．．．．．．．．．．．． 20
3.3 異質屬性．．．．．．．．．．．．．．．．．．．．．． 20
3.4 加權機制．．．．．．．．．．．．．．．．．．．．．． 22
3.4.1 正規化(Normalization) ．．．．．．．．．．．．．．． 22
3.4.2 資訊獲得值(IG) ．．．．．．．．．．．．．．．．．． 22
3.4.3 倒傳遞類神經網絡的貢獻度分析(CBP) ．．．．．．．．． 23
3.4.4 倒傳遞類神經網絡的鍊結權重(WBP) ．．．．．．．．．． 24
3.4.5 適應共振理論的鍊結權重(ART) ．．．．．．．．．．．． 25
3.4.6 基因演算法的染色體組合．．．．．．．．．．．．．．．． 25
3.4.7 多變量分析的變量權重．．．．．．．．．．．．．．．． 25
3.4.8 非線型研究的屬性權重(NP) ．．．．．．．．．．．．． 26
3.5 屬性萃取．．．．．．．．．．．．．．．．．．．．．． 26
3.6 相似度計算．．．．．．．．．．．．．．．．．．．．． 27
3.6.1 皮爾森法．．．．．．．．．．．．．．．．．．．．．． 27
3.6.2 向量法．．．．．．．．．．．．．．．．．．．．．．． 28
3.6.3 距離法．．．．．．．．．．．．．．．．．．．．．．． 28
3.7 分群法應用．．．．．．．．．．．．．．．．．．．．． 30
四、 實驗設計．．．．．．．．．．．．．．．．．．．．．． 31
4.1 研究架構．．．．．．．．．．．．．．．．．．．．．． 32
4.2 資料來源．．．．．．．．．．．．．．．．．．．．．． 34
4.3 實驗步驟設計．．．．．．．．．．．．．．．．．．．． 36
4.3.1 欄位初步選取．．．．．．．．．．．．．．．．．．．． 36
4.3.2 分群法選取．．．．．．．．．．．．．．．．．．．．． 36
4.3.3 屬性加權方法的使用．．．．．．．．．．．．．．．．． 37
4.3.4 相似度計算．．．．．．．．．．．．．．．．．．．．． 39
4.3.5 推薦及預測機制．．．．．．．．．．．．．．．．．．． 39
五、 實驗過程與結果分析．．．．．．．．．．．．．．．．． 40
5.1 實驗系統架構及說明．．．．．．．．．．．．．．．．． 40
5.2 資料前置處理．．．．．．．．．．．．．．．．．．．． 41
5.2.1 資料淨化(Data Cleaning) ．．．．．．．．．．．．．． 41
5.2.2 屬性選取(Feature Selecting) ．．．．．．．．．．．． 42
5.2.3 資料統整(Data Integration) ．．．．．．．．．．．． 42
5.2.4 資料轉換(Data Transformation) ．．．．．．．．．．． 43
5.2.5 資料離散化(Data Discretization) ．．．．．．．．．． 43
5.3 分群機制．．．．．．．．．．．．．．．．．．．．．． 43
5.3.1 正規化(Normalization) ．．．．．．．．．．．．．．． 43
5.3.2 資訊獲得值(IG) ．．．．．．．．．．．．．．．．．．． 44
5.3.3 倒傳遞類神經網絡的隱藏層節點貢獻度(CBP) ．．．．．． 45
5.3.4 倒傳遞類神經網路鍊結權重(WBP) ．．．．．．．．．． 47
5.4 綜合推薦機制 48
5.5 測試模組．．．．．．．．．．．．．．．．．．．．．． 49
5.5.1 測試指標．．．．．．．．．．．．．．．．．．．．．． 49
5.5.2 測試步驟．．．．．．．．．．．．．．．．．．．．．． 50
5.6 知識呈現模組．．．．．．．．．．．．．．．．．．．． 51
5.7 實驗一之結果分析．．．．．．．．．．．．．．．．．． 52
5.8 實驗二之結果分析．．．．．．．．．．．．．．．．．． 54
5.9 實驗三之結果分析 56
5.10 綜合討論．．．．．．．．．．．．．．．．．．．．．． 57
六、 結論與建議．．．．．．．．．．．．．．．．．．．．． 59
6.1 研究結論．．．．．．．．．．．．．．．．．．．．．． 59
6.2 未來研究之建議．．．．．．．．．．．．．．．．．．． 61
參考文獻．．．．．．．．．．．．．．．．．．．．．． 63
附件一 Data Mining運用的理論與實際應用功能．．．．．．．． 67
附件二 資料字典．．．．．．．．．．．．．．．．．．．．．． 69
附件三 資料前置處理篩選規則．．．．．．．．．．．．．．．． 71

[01] Robert E. Wayland and Paul M. Cole合著, “CUSTOMER CONNECTIONS-- New Strategies For Growth”, 邱振儒譯，客戶關係管理─創造企業與客戶重複互動的客戶聯結技術，商周出版，台北，民國八十八年。
[02] David E. Bell and Arthur Schieifer Jr.合著, “Decision Making Under Uncertainty”, 李茂興，劉原彰合譯，不確定情況下的決策，王秉鈞主編，弘智出版，台北，民國八十八年。
[03] Alex Berson, Stephen Smith and Kurt Thearling合著, “Building Data Mining Applications for CRM”, 葉涼川譯，CRM Data Mining應用系統建置，麥格羅．希爾出版，台北，民國九十年。
[04] Ronald S. Swift著, “ACCELERATING Customer Relationships-- Using CRM and Relationship Technologies”, 賴士奇，吳嘉哲，劉揚愷，楊子蕙合譯，深化顧客關係管理，遠擎出版，台北，民國九十年。
[05] Rebort S. Kaplan and David P. Norton合著, “THE BALANCED SCORECARD-- Translating Strategy into Action”, 朱道凱譯，平衡計分卡，臉譜出版，台北，民國八十八年。
[06] Paul R. Niven著, “BALANCED SCORECARD STEP BY STEP─Maximizing Performance and Maintaining Results”, 于泳泓譯，平衡計分卡最佳實務，商周出版，台北，民國九十一年。
[07] 蘇木春、張孝德編著，機器學習 類神經網路、模糊系統以及基因演算法則，全華出版、台北、民國八十九年。
[08] Jiawei Han and Micheline Kamber, "Data Mining: Concepts and Techniques", The Morgan Kaufmann Series in Data Management Systems, Jim Gray, Series Editor Morgan Kaufmann Publishers, August 2000.
[09] “ISO/IEC 13250,Topic Maps”, Information Technology Document Description and Processing Languages, 13250, First, December 3, 1999,E
[10] Alex Berson and Stephen J. Smith, “Data Warehousing, Data Mining & OLAP”, McGraw-Hill , New York, 1997.
[11] A.K. Jain, M.M. Murty and P.J. Flynn: “Data Clustering: A Review”, ACM Computing Surveys, Vol. 31, No. 3, September 1999.
[12] J.L. Gorden, “Creating knowledge maps by exploiting dependent relationships”, Knowledge-Base Systems, 13, pages 71-79, 2000.
[13] Kobus Barnard, Pinar Duygulu and David Forsyth, “Clustering Art”, Proceedings of the 2001 IEEE Computer Society Conference on , Volume: 2 , Pages II-434 -II-441 vol.2, 2001.
[14] Aggarwal, C.C. and Yu, P.S.,” Redefining clustering for high-dimensional applications”, Knowledge and Data Engineering, IEEE Transactions on , Volume: 14 Issue: 2 , Page(s): 210 -225, Mar/Apr 2002.
[15] Yeung, D.S. and Wang, X.Z.,” Improving performance of similarity-based clustering by feature weight learning”, Pattern Analysis and Machine Intelligence, IEEE Transactions on , Volume: 24 Issue: 4 , Page(s): 556 -561, Apr 2002.
[16] Li, C. and Biswas, G., “Unsupervised learning with mixed numeric and nominal data”, Knowledge and Data Engineering, IEEE Transactions on , Volume: 14 Issue: 4 , Page(s): 673 -690, Jul/Aug 2002.
[17] Daxin Jiang, Jian Pei, Aidong Zhang, “DHC:a density-based hierarchical clustering method for time series gene expression data”, Daxin Jiang; Jian Pei; Aidong Zhang; Bioinformatics and Bioengineering, 2003. Proceedings. Third IEEE Symposium on , Page(s): 393 -400, 2003.
[18] Gholamhosein Shekholeslami, Wendy Chang and Aidong Zhang: “SemQuery: Semantic Clustering and Querying on Heterogeneous Features for Visual Data”, IEEE Trans. On Knowledge and data engineering, Vol. 14, No. 5, Sep/Oct 2002.
[19] C. H. Yun, K. T. Chuang and M. S. Chen: "An Efficient Clustering for Market Basket Data Based on Small Large Ratios", COMPSAC 2001.
[20] W. Li, J Han and J Pei: "CMAR: Accurate and Efficient Classification Based on Multiple Class-Association Rules", NSERC-A3723, NCE/IRIS-3.
[21] Fang-Xiang Wu, Wen-Jun Zhang and Kusalik, A.J., “Determination of the minimum sample size in microarray experiments to cluster genes using k-means clustering”, Bioinformatics and Bioengineering, 2003. Proceedings. Third IEEE Symposium on , Page(s): 401 -406, 2003.
[22] Erdoan, H., Yuqing Gao and Picheny, M., ” Rapid adaptation using penalized-likelihood methods” ,Acoustics, Speech, and Signal Processing, 2001. Proceedings. (ICASSP ''01). 2001 IEEE International Conference on , Volume: 1 , Page(s): 333 -336 vol.1, 2001.
[23] Fei Wu and Gardarin, G.,” Gradual clustering algorithms” ,Database Systems for Advanced Applications, 2001. Proceedings. Seventh International Conference on , Page(s): 48 -55, 2001.
[24] Duntong Yu; Aidong Zhang; ,” ClusterTree: integration of cluster representation and nearest neighbor search for image databases” ,Multimedia and Expo, 2000. ICME 2000. 2000 IEEE International Conference on , Volume: 3 , Page(s): 1713 -1716 vol.3, 2000.
[25] Shouhong Wang and Hai Wang, “Knowledge Discovery Through Self-Organizing Maps: Data Visualization and Query Processing”, Knowledge and Information Systems, (2002) 4: 31-45.
[26] Stefan Smolnik, Ludwig Nastansky, "K-Discovery: Identification of Distributed Knowledge Structures in a Process-Oriented Groupware Environment", 2001.
[27] Beckenkamp, F.G., Feldens, M.A. and Pree, W., ” Optimization of the combinatorial neural model ”,Neural Networks, 1998. Proceedings. Vth Brazilian Symposium on , Page(s): 49 -54, 9-11 Dec 1998.
[28] Chun-Hung Cheng, Ada Waichee Fu and Yi Zhang, “Entropy-based subspace clustering for mining numerical data”, Proceedings of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining, San Diego, California, United States, Pages: 84 — 93, 1999.
[29] Yan Huang, Hui Xiong, Shashi Shekhar and Jian Pei, “Mining Confident Co-location Rules without A support Threshold”, CAN SAC Melbourne, Florida, USA, 2002.
[30] SAS Institute Inc.: "Analyzing the footsteps of your customers", SAS Reach Report, 2000.
[31] Fausett, L., “Fundamentals of Neural Networks” , Prentice-Hall International, 1994.
[32] Patterson, D. W., “Artiticial Neural Networks: Theory and Applications”, Prentice Hall, 1996.
[33] Haykin, S., “Neural Networks: A Comprehensive Foundation”, Prentice Hall, 1999.
[34] Hecht-Nielsen, R., “Neurocomputing”, Addison-Wesley, 1991.
[35] Feldkamp, L.A., Prokhorov, D.V. and Feldkamp, T.M., “Cluster-weighted modeling with multiclusters”, Neural Networks, 2001. Proceedings. IJCNN ''01. International Joint Conference on , Volume: 3 ,Page(s): 1710 -1714 vol.3, 2001.
[36] Prokhorov, D.V., Feldkamp, L.A. and Feldkamp, T.M., “A new approach to cluster-weighted modeling”, Neural Networks, 2001. Proceedings. IJCNN ''01. International Joint Conference on , Volume: 3 , Page(s): 1669 -1674 vol.3, 2001.
[37] Tsang, E.C.C., Shiu, S.C.K., Wang, X.Z. and Lam, M.; “Clustering and classification of cases using learned global feature weights”, IFSA World Congress and 20th NAFIPS International Conference, 2001. Joint 9th , Page(s): 2971 -2976 vol.5, 25-28 July 2001.
[38] Erdem, C. and Zimmermann, H.G., “A data-driven method for input feature selection within neural prosody generation”, Acoustics, Speech, and Signal Processing, 2002. Proceedings. (ICASSP ''02). IEEE International Conference on , Volume: 1 , Page(s): I-477 -I-480 vol.1, 2002.
[39] Tsang, E.C.C., Shiu, S.C.K., Wang, X.Z. and Ho, K., “Case-base reduction using learned local feature weights”, IFSA World Congress and 20th NAFIPS International Conference, 2001. Joint 9th , Page(s): 2965 -2970 vol.5, 25-28 July 2001.
[40] Paul Resnick, Neophytos Iacovou, Mitesh Suchak, Peter Bergstrom and John Riedl, “GroupLens: An Open Architecture for Collaborative Filtering of Netnews”, Proc. Of ACM 1994, Conf. on Computer Supported Cooperative Work, Chapel Hill, Pages 175-186.
[41] Paul Resnick and Hal R. Varian, “Recommender systems ”, Communications of the ACM , Volume 40 , Issue 3 (March 1997), Pages: 56 — 58, 1997.
[42] Loren Terveen, Will Hill, Brian Amento,David McDonald and Josh Creter, ” PHOAKS: a system for sharing recommendations”, Communications of the ACM, Volume 40 , Issue 3 (March 1997), Pages: 59 — 62,1997.
[43]Wolfgang Banzhaf, “Foundations of Genetic Algorithms”, et al, 1999.