臺灣博碩士論文加值系統

過去學者提出的推薦研究裡，有分析個人資料檔（User Profile）找出其喜好項目來進行推薦，或從具相似喜好的顧客中找出排行榜（Top-N）以做為推薦項目，近來更常見推薦系統加入RFM分析法分析顧客的消費特性，即利用R（Recency）、F（Frequency）及M（Monetary）三個指標來衡量顧客的忠誠度。然而過去學者所提出RFM結合推薦系統的相關研究中，常忽略產品特性與購買週期之間的關係，因此隱含了些許的缺失，例如計算R值時常採用固定區間長度的切割來給予不同的分數，但顧客對於某些產品的購買行為是有週期性的，所以會產生推薦時機不適當的問題，即未考慮產品適時性。因此，不論在內容導向式推薦、協同過濾式推薦或RFM結合推薦機制的方法中，都未考慮顧客在不同時間對不同產品會有重複購買的需求，當顧客在進行消費後，隨著產品的消耗或顧客購買週期的產生，都可能會引起顧客更強烈的需求，但是以往的推薦系統在產生推薦項目時，常因顧客對產品需求的週期未達到而造成無效的推薦，因而降低推薦效率。
針對以上的問題，本研究提出一個基於RFM分析法之適時性推薦機制（Timely RFM，TRFM），本方法結合了「產品適時性」及「顧客消費特性」，並利用類神經網路的自適應共振理論網路，有效的將購買行為相似的顧客做快速的群聚，用以分析顧客對企業的忠誠度及貢獻度，並可同時了解顧客的消費行為，以達到更有效的分群結果。本研究所達成之目標為(1)分析不同產品對每位顧客在不同時間的需求狀況，(2)提供顧客一個適時、適性的推薦機制，(3)改善現有RFM結合推薦系統方法的不足，使量化後的顧客購買行為更具代表性。最後，本研究在實驗中證明，加入產品週期性概念可提供顧客一個適時性的產品推薦。

The recent study of recommendation systems and RFM method has been applied to analyze customers’ consumption property and the re-purchasing ability. The RFM method employs Recency (R), Frequency (F), and Monetary (M) to measure customers’ consumption loyalty. The recommendation systems are mainly to promote products for increasing profit. However, there are some problems because these approaches ignore the relationship between product property and purchase periodicity. That is, the combination of recommendation systems and RFM method did not take the customers product-purchasing timing into consideration. If the periodicity of product-demand can be estimated by each customer’s buying behavior, then the product recommendation at the right timing shall match the buying requirement. This is the reason why the past product recommendation studies have difficulty of increasing the accuracy. To deal with the product periodicity, this research proposes a Timely RFM (TRFM) method which takes product property and purchase periodicity into consideration. This method uses Adaptive Resonance Theory (ART) of Artificial Neural Network (ANN) to cluster customers based on their purchasing behavior in order to obtain similar interest of customer. This research is intended (1) to analyze different products to each customer’s demands in different times, (2) to provide a recommendation mechanism to satisfy customers’ needs, and (3) to improve the deficiency of existing combination with recommendation and RFM. To examine the practicability and to validate the method, the experimentation uses the Foodmart2000 database of Microsoft SQL2000 to verify the accuracy of TRFM. The results prove that our proposed method can provide a timely recommendation and create better results than non-timely recommendation.

摘要 I
Abstract II
誌謝 III
目錄 V
表目錄 IX
圖目錄 XI
第一章 緒論 1
1.1 研究背景 1
1.2 研究動機與目的 4
1.3 研究步驟 5
1.4 研究範圍 6
1.5 論文架構 6
第二章 文獻探討 8
2.1 推薦系統 8
2.1.1 內容導向式推薦 9
2.1.2 協同過濾式推薦 11
2.1.3 混合式推薦 15
2.2 市場區隔與RFM分析法 16
2.2.1 市場區隔 16
2.2.2 RFM分析法 16
2.2.3 顧客五等分法（Customer Quintiles Method） 18
2.3 類神經網路（Artificial Neural Network） 18
2.3.1 類神經網路型態 19
2.3.2 自適應共振理論網路（Adaptive Resonance Theory） 20
2.3.3 運用膝點決定群數 25
2.4 推薦效能評估方法 26
第三章 適時性推薦機制（TRFM） 28
3.1 TRFM架構及處理步驟 29
3.2 資料蒐集及預處理 33
3.2.1 資料蒐集 33
3.2.2 資料預處理（Data Preprocessing） 34
3.3 RFM分析模組（RFM Analysis） 36
3.4 群聚模組（Clustering Process） 42
3.4.1 資料轉換 44
3.4.2 產生輸入資料集合 44
3.4.3 群聚結果 45
3.5 適時性評估模組（Timely Estimation） 46
3.6 推薦模組 47
3.6.1 Low-RFM之推薦階段 47
3.6.2 High-RFM之推薦階段 49
3.7 推薦效能評估 51
第四章 實驗分析 52
4.1 系統環境及使用工具 52
4.2 資料來源 53
4.3 實驗設計 54
4.4 顧客交易資料分析 55
4.5 RFM初始分數分析 57
4.6 RFM五等分結果 59
4.7 ART分群 60
4.7.1 資料設計及編碼 61
4.7.2 警戒值設定 62
4.7.3 分群結果 63
4.7.4 產生Low-RFM顧客群之推薦列表 63
4.8 適時性產品分析 64
4.8.1 顧客之產品購買週期分析 65
4.8.2 產生High-RFM顧客群之推薦列表 65
4.9 推薦預測效能評估 66
第五章 結論與未來研究 69
5.1 結論與研究貢獻 69
5.2 未來研究 70
參考文獻 71
附錄一 自適應共振理論網路分群結果 76

[1]吳志宏(2004)，以隱性回饋為基礎的自動化推薦機制，碩士論文，朝陽科技大學資訊管理系，台中。
[2]李麗華、李富民、黃鈁媖(2005)，「運用類神經網路預測新進顧客產品喜好之個人化商品推薦技術」，2005電子商務與數位生活研討會，台北，第1116-1130頁。
[3]邱宏彬、蘇建源(2004)，「一個可彈性支援顧客關係管理與資料庫行銷之模糊RFM Model」，電子商務學報，第六卷，第二期，第149-173頁。
[4]郭盛揚(2004)，以顧客交易特徵為基礎的顧客分群方法，碩士論文，朝陽科技大學資訊管理系，台中。
[5]葉怡成(1999)，類神經網路模式應用與實作，儒林圖書公司。
[6]榮泰生(2000)，網路行銷電子商務實務篇，五南圖書出版有限公司。
[7]劉文良(2004)，網路行銷理論與實務，金禾資訊股份有限公司。
[8]Chesani, Federico(2002), “Recommendation Systems,” Corso di laurea in Ingegneria Informatica.
[9]Cheung, Kwok-Wai, Kwok, James T., Law, Martin H., and Tsui, Kwok-Ching(2003), “Mining customer product ratings for personalized marketing,” Decision Support Systems, Vol. 35, No. 2, pp. 231-243.
[10]Cho, Ho Yoon and Kim, Jae Kyeong (2004), “Application of Web usage mining and product taxonomy to collaborative recommendations in e-commerce,” Expert Systems with Applications, Vol. 26, No. 2, pp. 233-246.
[11]Cho, Yeong Bin, Cho, Yoon Ho, and Kim, Soung Hie(2005), “Mining changes in customer buying behavior for collaborative recommendations,” Expert Systems with Applications, Vol. 28, No. 2, pp. 359-369.
[12]Deshpande, Mukund and Karypis, George(2004), “Item-Based Top-N Recommendation Algorithms,” ACM Transactions on Information Systems, Vol. 22, No. 1, pp. 143-177.
[13]G.A. Carpenter and S. Grossberg(1986), “Neural dynamics of category learning and recognition: Attention, memory consolidation, and amnesia,” In J. Davis, R. Newburgh, and E. Wegman, editors, Brain Structure, Learning, and Memory,” AAAS Symposium Series, pp. 239-286.
[14]Goldberg, David, Nichols, David, Oki, Brian M., and Terry, Douglas(1992), “Using collaborative Filtering to Weave an Information TAPESTRY,” Communications of the ACM, Vol. 35, No. 12, pp. 61-70.
[15]Han, Peng, Xie, Bo, Yang, Fan, and Shen, Ruimin(2004), “A scalable P2P recommender system based on distributed collaborative filtering,” Expert Systems with Applications, Vol. 27, No. 2, pp. 203-210.
[16]Kim, B-D. and Kim, S-O.(2001), “A new recommender system to combine content-based and collaborative filtering systems,” The Journal of Database Marketing, Vol. 8, No. 3, pp. 244-252.
[17]Kim, Won, Ko, Il-Ju, Yoon, Jin-Sung, and Kim, Gye-Young(2004), “Inference of recommendation information on the internet using improved FAM,” Future Generation Computer Systems, Vol. 20, No. 2, pp. 265-273.
[18]Lee, D.H., Kim, S.H., and Ahn, B.S.(2000), “A Conjoint Model for Internet Shopping Malls Using Customer''s Purchasing Data,” Expert Systems with Applications ,Vol. 19, No. 1, pp. 59-66.
[19]Li, Yu, Lu, Lin, and Xuefeng, Li(2005), “A hybrid collaborative filtering method for multiple-interests and multiple-content recommendation in E-Commerce,” Expert Systems with Applications, Vol. 28, No.1, pp. 67-77.
[20]Liu, Duen-Ren and Shih, Ya-Yueh(2005), “Hybrid approaches to product recommendation based on customer lifetime value and purchase preferences,” The Journal of Systems and Software, Vol. 77, No. 2, pp.181-191.
[21]Liu, Duen-Ren and Shih, Ya-Yueh(2005), “Integrating AHP and data mining for product recommendation based on customer lifetime value,” Information and Management, Vol. 42, No. 3, pp. 387-400.
[22]Miglautsch(2000), John R., “Thoughts on RFM Scoring,” Journal of Database Marketing, Vol. 8, No. 1, pp. 67-72.
[23]Mild, Andreas and Reutterer, Thomas(2003), “An improved collaborative filtering approach for predicting cross-category purchased based on binary market basket data,” Journal of Retailing and Consumer Services, Vol. 10, No. 3, pp. 123-133.
[24]Min, Sung-Hwan and Han, Ingoo(2005), “Detection of the customer time-variant pattern for improving recommender systems,” Expert Systems with Applications, Vol. 28, No. 2, pp. 189-199.
[25]Muammer, Ozer(2001), “User Segmentation of Online Music Services Using Fuzzy Clustering,” Omega, Vol. 29, No. 2, pp. 193-206.
[26]Resnick, Paul and Varian, Hal R.(1997), “Recommender Systems,” Communications of the ACM, Vol. 40, No. 3, pp.56-58.
[27]Roh, Tae Hyup, Oh, Kyong Joo, and Han, Ingoo(2003) “The collaborative filtering recommendation based on SOM cluster-indexing CBR,” Expert Systems with Applications, Vol. 25, No. 3, pp. 413-423.
[28]Sarwar, Badrul, Karypis, George, Konstan, Joseph, and Riedl, John(2000), “Analysis of Recommendation Algorithms for E-commerce,” Proceedings of the 2nd ACM Conference on Electronic Commerce, pp.158-167.
[29]Schafer, J. Ben, Konstan, Joseph A., and Riedl, John(2001), “E-Commerce Recommendation Applications,” Data Mining and Knowledge Discovery, Vol. 5, No. 1, pp.115-153.
[30]Schafer, J. Ben, Konstan, Joseph, and Riedl, John(1999), “Recommender Systems in E-commerce,” Proceedings of the First ACM Conference on Electronic Commerce, pp.158-166.
[31]Senecal, Sylvain, Kalczynski, Pawel J., and Nantel, Jacques(2005), “Consumers’ decision-making process and their online shopping behavior: a clickstream analysis,” Journal of Business Research, Vol. 58, No. 11, pp. 1599-1608.
[32]Senecal, Sylvain and Nantel, Jacques(2004), “The Influence of Online Recommendation on Consumer’s Choices,” Journal of Retailing, Vol. 80, No. 2, pp.159-169.
[33]Shardanand, Upendra and Maes, Pattie(1995), “Social Information Filtering: Algorithms for Automating “Work of Mouth”,” Proceedings of the SIGCHI conference on Human factors in computing systems, pp.210-217.
[34]Stone Bob (1989), Successful Direct Marketing Methods, 4th ed., NTC Business Books.
[35]Suh, E.H., Noh, K.C., and Suh, C.K.(1999), “Customer list segmentation using the combined response model,” Expert Systems with Applications, Vol. 17, No. 2, pp.89-97.
[36]Sung Ho Ha and Sang Chan Park(1998), “Application of data mining tools to hotel data mart on the Internet for database marketing,” Expert Systems with Application, Vol. 15, pp. 1-31.
[37]Tsai, C.-Y. and Chiu, C.-C.(2004), “A purchase-based market segmentation methodology,” Expert Systems with Applications, Vol. 27, No. 2, pp. 265-276.
[38]Wang, Feng-Hsu and Shao, Hsiu-Mei(2004), “Effective personalized recommendation based on time-framed navigation clustering and association mining,” Expert Systems with Applications, Vol. 27, No. 3, pp. 365-377.
[39]Weng, Sung-Shun and Liu, Mei-Ju(2004), “Feature-based recommendation for one-to-one marketing,” Expert Systems with Applications, Vol. 26, No. 4, pp.493-508.
[40]Yuan, Soe-Tsyr and Tsao, Y. W.(2003), “A recommendation mechanism for contextualized mobile advertising,” Expert Systems with Applications, Vol. 24, No. 4, pp. 399-414.