(34.204.201.220) 您好!臺灣時間:2021/04/20 12:27
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:黃鈁媖
研究生(外文):Fang-Ying Huang
論文名稱: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
論文名稱(外文):Using Neural Network to Predict New Customer''s Preference for Better Recommendation
指導教授:李麗華李麗華引用關係
指導教授(外文):Li-Hua Li
學位類別:碩士
校院名稱:朝陽科技大學
系所名稱:資訊管理系碩士班
學門:電算機學門
學類:電算機一般學類
論文種類:學術論文
論文出版年:2005
畢業學年度:94
語文別:中文
論文頁數:102
中文關鍵詞:Fuzzy RFMRFM分析法個人化行銷推薦系統倒傳遞網路
外文關鍵詞:RFMOne-to-one MarketingRecommender SystemsBack-Propagation Network (BPN)Fuzzy RFM
相關次數:
  • 被引用被引用:9
  • 點閱點閱:1296
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:264
  • 收藏至我的研究室書目清單書目收藏:5
推薦系統用於推薦顧客商品喜好已行之有年,其中協同式過濾方法(Collaborative filtering)的推薦技術,有稀疏性(Sparsity)、延展性(Scalability)及冷啟始(Cold-start)等問題,目前已有眾多學者提出方法解決稀疏性和延展性問題,但冷啟始問題卻甚少學者提出解決之道,冷啟始問題即因為新進顧客交易量少資訊不足而在推薦上有較大的困難,然而網路使用者來來去去,故如何預測新顧客之喜好來推薦產品愈來愈重要。有鑑於此,本研究針對此問題提出一個利用RFM分析法的最近購買時間(Recency, R)、購買次數(Frequency, F)及購買金額(Monetary, M)等三項資料將之語意化來對顧客分群,搭配倒傳遞網路(BPN)來預測新顧客所喜好的個人化商品,最後利用Top-N做推薦,如此可以達到本研究欲達成的目標:(1)減少冷啟始問題及(2)利用衡量指標計算推薦效果,並分析R、F及M三個指標何者較為重要,以使推薦更加準確。
本研究利用一個大型賣場之交易資料庫為實作樣本,利用本研究所提方法分析RFM三指標權重不同而分組分析,實驗結果顯示在利用Top-N推薦產品給新進顧客的推薦成效上,Precision值最高可達100%,而F1值也高達74%,而且由本實驗結果得知是在推薦顧客購買六項以上商品給新進顧客會有不錯的成效,所以商家可依此對各種類型的顧客群找出更合適的行銷方法;而實驗結果亦顯示R即購買期間與M即花費經額指標對該賣場較為重要。以上結果可看出本研究方法在推薦預測上是有效的推薦技術,同時亦顯示本研究所提的方法能降低冷啟始的問題。
關鍵字:倒傳遞網路(Back-Propagation Network , BPN)、
推薦系統、個人化行銷、RFM分析法、Fuzzy RFM。
Abstract
Web recommendation systems have been applied to recommend products based on customers’ preference. There are many techniques being used for recommendation such as collaborative filtering. However, there are also exists some problems such as sparsity of transaction records, scalability of recommendation mode, and cold-start problem. To solve sparsity and scalability problems, many techniques have been proposed, however, cold start problem is rarely studied. Cold-start problem usually means the difficulty of providing useful information to new customer because of the insufficient information about the customer.
Due to the booming of web society, new users are growing every day, let alone the net surfing behavior creates variety of choice. Therefore, how to service new users and to recommend the right products becomes very important. This is the reason of this research to deal with cold-start problem to provide better recommendation.
The research is aimed at the user preference and to predict the right product for the new user. The first step is to cluster customers by providing semantic meaning based on recency, frequency, and monetary. The second step is to employ BPN to predict new customers’ preferred merchandise. Finally, the Top-N products are suggested for recommendation. The objectives of this thesis are:(1) to solve the problem of cold start and (2) to determine which parameter is more important with R, F, M indicators.
The implementations of the method in this thesis are described as follows. The data source is collected from a supermarket. Fuzzy RFM is used in this thesis and its related weights are employed for cluster analysis. The experimental result shows that precision measure can reach 100% and F1 measure can be at most 74% with respect to the performance of recommending Top-N merchandise to new customers. Furthermore, this experiment performs effectively while recommending new customers more than six products. And, the firms can find out more appropriate marketing strategies for a variety of customers. Also, our experiment shows that indicators R and M are more important to this specific supermarket. Considering the effectiveness shown in the experiments, our methods can be a good recommendation technique in terms of the effectiveness and can also solve the problem of cold start.
Keyword: Back-Propagation Network (BPN), Recommender Systems, One-to-one Marketing , RFM , Fuzzy RFM.
目 錄
摘   要 I
AbstractIII
誌 謝 V
目 錄 VII
圖 目 錄 X
表 目 錄 XI


第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的 8
1.3 研究步驟 9
1.4 論文架構 11
第二章 文獻探討 13
2.1 個人化行銷 13
2.2 推薦系統 14
2.2.1推薦系統的定義及特性 14
2.2.2推薦系統技術介紹 15
2.2.2.1 內容式導向(Content-Based, CB) 16
2.2.2.2 協同式過濾(Collaborative Filtering, CF) 18
2.2.2.3 混合式過濾(Hybrid Approaches) 25
2.2.3推薦系統形式種類 26
2.3市場區隔與市場區隔分析法 27
2.3.1 市場區隔 28
2.3.2 市場區隔的區隔類型 29
2.3.3市場區隔變數 30
2.3.4市場區隔分析法 32
2.3.5 RFM分析法與Fuzzy RFM分析法 34
2.4類神經網路(Artificial Neural Network, ANN) 39
2.4.1 類神經網路的定義 39
2.4.2 類神經網路的網路型態 40
2.4.3 倒傳遞網路(BPN) 41
2.5效能評估 45
2.5.1回應率(Recall, R) 46
2.5.2準確率(Precision, P) 47
2.5.3 F1指標 47
第三章 研究方法 48
3.1 研究架構 48
3.2 方法與步驟 52
3.3 推薦模組設定 55
3.3.1 初始分數設定 55
3.3.2 推薦分數設定 64
3.3.3 倒傳遞網路輸入輸出值設定 66
3.4 效能評估指標 67
第四章 實驗分析 70
4.1 實驗資料來源 70
4.2 資料環境設定 71
4.3 顧客屬性分析 71
4.4 實驗分組分析 72
4.5 實驗步驟分析 75
4.6 實驗結果分析 83
第五章 總結與未來研究 93
5.1 總結 93
5.2 研究貢獻 93
5.3 未來研究方向 94
參考文獻 96


圖 目 錄
圖一 研究步驟流程圖 11
圖二 購買行為模式矩陣 19
圖三 推薦系統組成架構 20
圖四 行銷策略的核心 29
圖五 倒傳遞網路架構圖 42
圖六 倒傳遞網路架構圖常用的非線性轉換函數 43
圖七 本研究架構 49
圖八 本研究流程圖 51
圖九 推薦模組細項流程 52
圖十 (a)三角隸屬函數圖 (b)梯形隸屬函數圖 60
圖十一 本研究之BPN網路架構示意圖 81
圖十二 實驗一:F1效能指標值比較分析圖 87
圖十三 實驗二:F、M權重比例不同的F1指標值比較分析圖 89
圖十四 實驗三:F1效能指標值比較分析圖 91
圖十五 三項實驗的F1效能指標值比較分析圖 92


表 目 錄
表一 行銷活動演進過程 1
表二 資訊擷取與資訊過濾之比較 3
表三 推薦方法的優缺點 26
表四 區隔變數之項目 32
表五 RFM分析法相關研究整理 36
表六 各學者運用RFM分析法之問題整理 38
表七 倒傳遞網路模式之整理 45
表八 衡量指標之變數示意表 46
表九 本研究初始分數變數符號說明 57
表十 顧客購買商品資訊範例 62
表十一 實驗二:FM權重不同之分組人數資料表 73
表十二 實驗二:FM權重不同之分組人數資料表 74
表十三 實驗三:RFM權重不同之分組人數資料表 75
表十四 產品類別名稱總觀 76
表十五 推薦分數(列舉) 77
表十六 顧客屬性名稱與倒傳遞網路輸入設定值轉換 78
表十七 顧客屬性正規化後的倒傳遞輸入串列表(列舉) 79
表十八 倒傳遞網路的輸出串列表(列舉) 80
表十九 倒傳遞網路中的新進顧客輸入串列表(列舉) 82
表二十 推薦列表(列舉) 82
表二十一 實驗一:R、F、M權重比例不同之Recall指標值比較表 85
表二十二 實驗一:R、F、M權重比例不同之Precision指標比較表 85
表二十三 實驗一:R、F、M權重比例不同之 F1指標值比較表 86
表二十四 實驗二:F、M權重比例不同的 Precision指標值比較表 87
表二十五 實驗二:F、M權重比例不同的Recall指標值比較表 88
表二十六 實驗二:F、M權重比例不同的F1指標值比較表 89
表二十七 實驗三:購買六項產品以上的效能指標分析表 91
[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學系,台中。
[2]李旭東 (2000),行銷學精要,高立圖書有限公司。
[3]邱永祥(2003),運用類神經網路與資料探勘技術於網路教學課程推薦之研究,碩士論文,朝陽科技大學資訊管理學系,台中。
[4]邱宏彬與蘇建源(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,Vol.6,No.2,pp.149-173。
[5]林娟如 (2000), 資訊時代的電子行銷, 研考雙月刊,Vol.24, No.3,pp.33-40。
[6]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研究系,台中。
[7]許明風(2001),個人化行銷決策系統之研究,碩士論文,朝陽科技大學資訊管理研究系,台中。
[8]葉怡成(1999),類神經網路模式應用與實作,儒林圖書公司。
[9]廖虹雲(2004),利用資料探勘來預測顧客對不同產品類別之偏好程度,碩士論文,朝陽科技大學資訊管理研究系,台中。
[10]廖婉菁(2002),應用協同過濾機制於商品推薦之研究-以手機網站為例,碩士論文,中原大學資訊管理學系,中壢。
[11]S. Alfred (1981), “Market Segmentation by Personal values and Salient Product Attributes,” Journal of Advertising Research, Vol.21, No.1, pp. 29-35.
[12]C. L. Bauer (1988), “A Direct Mail Customer Purchase Model,” Journal of Direct Marketing, Vol. 2, pp. 16-24.
[13]S. Wesley Changchien and Tzu-Chuen Lu (2002), “Mining Association Rules Procedure to Support On-line Recommendation by Customers and Products Fragmentation,” Expert Systems with Applications, Vol. 20, No. 4, pp. 325-335.
[14]Kwok-Wai Cheung, James T. Kwok, Martin H. Law ,and Kwok-Ching Tsui (2003), “Mining Customer Product Ratings for Personalized Marketing,” Decision Support Systems - special issue on Web Data Mining, Vol. 35, No. 2, pp. 231-243.
[15]Yoon Ho Choa and Jae Kyeong Kim (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.
[16]Federico Chesani (2002), Recommendation Systems, matr. 2148054547, Corso di laurea in Ingegneria Information.
[17]David Goldberg, David Nichols, Brian M. Oki ,and Douglas Terry (1992), “Using Collaborative Filtering to Weave an Information TAPESTRY,” Communications of the ACM, Vol. 35, No. 12, pp. 61-70.
[18]Peng Han, Bo Xie, Fan Yang ,and Ruimin Shen (2004) , “A Scalable P2P Recommender System based on Distributed Collaborative Filtering,” Expert Systems with Applications, Vol. 27, No. 2, pp. 203-210.
[19]Harald Hruschka and Martin Natter (1999) , “Comparing Performance of Feedforward Neural Nets and K-means for Cluster-based Market Segmentation,” European Journal of Operational Research, Vol. 114, No. 21999, pp. 346-353.
[20]P. Kotler (1994), Marketing Management:Analysis, Planning, Implementation, 8th ed., Prentice Hall, Inc.
[21]P. Kotler and G. Armstrong (1997), Principles of Marketing, Prentice Hall.
[22]D.H. Lee, S.H. Kim ,and B.S. Ahn (2000), “A Conjoint Model for Internet Shopping Malls Using Customer''s Purchasing Data,” Expert Systems with Applications ,Vol. 19, No. 1, pp. 59-66.
[23]T. Levitt (1960), ”Marketing Mypia,” Harvard Business Review, pp. 45-56.
[24]P. Melville, Raymond J. Mooney ,and Ramadass Nagarajan (2001), “Content-Boosted Collaborative Filtering,” ACM GIR-2001 Workshop on Recommender Systems, New Orleans, L.A.
[25]Stuart E. Middleton, Harith Alani, Nigel R. Shadbolt ,and David C. De Roure(2002), ”Exploiting Synergy Between Ontologies and Recommender Systems”, In Proceedings of Semantic Web Workshop 2002 At the Eleventh International World Wide Web Conference Hawaii, USA.
[26]John Miglautsch (2000), “Thoughts on RFM Scoring”, Journal of Database Marketing, Vol.8, No.1.
[27]Ozer Muammer (2001), “User Segmentation of Online Music Services Using Fuzzy Clustering,” Omega, Vol. 29, No. 2, pp. 193-206.
[28]D.W. Oard and G. Marchionini (1996), “A Conceptual Framework for Text Filtering,” Technical Report EE-TR-96-25, CAR-TR-830, CLIS-TR-96-02, CS-TR-3643, University of Maryland, College Park.
[29]M.J. Pazzani (1999) , “A Framework For Collaborative, Content-Based and Demographic Filtering,” Artificial Intelligence Review, Vol. 13, No. 5-6 , pp.393-408.
[30]B. J. Pine, B. B. Victor, and C. Andrew (1993), "Making Mass Customizaion Work," Harvard Business Review, Vol.71, No.5, pp.108-111.
[31]Paul Resnick, Neophytos Iacovou, Mitesh Suchak, Peter Bergstrom ,and John Riedl (1994), “Group Lens:An Open Architecture for Collaborative Filtering of Netnews,” Proceedings of ACM CSCW’94,Conference on Computer Supported Cooperative Work, pp. 175-186.
[32]P. Resnick and H. R. Varian (1997), “Recommender Systems,” Communications of ACM, Vol. 40, No. 3, pp. 56-58.
[33]Tae Hyup Roh, Kyong Joo Oh ,and Ingoo Han (2003), “The Collaborative Filtering Recommendation based on SOM Custer-indexing CBR,” Expert Systems with Applications, Vol.25, No. 3, pp. 413-423.
[34]Badrul Sarwar, Karypis George, Joseph Konstan ,and John Riedl (2000), “Analysis of Recommendation Algorithms for E-commerce,” Proceedings of the 2nd ACM conference on Electronic commerce, pp.158-167.
[35]J. Ben Schafer, Joseph Konstan, and John Riedl (1999), “Recommender System in E-commerce,” Proceedings of the First ACM Conference on Electronic Commerce, pp.158-166.
[36]Andrew I. Schein, Alexandrin Popescul, Lyle H. Ungar ,and David M. Pennock (2002), “Collaborative Filtering: Methods and Metrics for Cold-Start Recommendations,” Annual ACM Conference on Research and Development in Information Retrieval, pp.253 – 260.
[37]D.C Schmittlein, L.G. Cooper ,and D.G. Morrison (1993), “Truth in Concentration in the Land of (80/20) Laws,” Marketing Science, Vol. 12, No. 2, pp. 167-183.
[38]Bob Stone (1989), Successful Direct Marketing Methods, 4th ed., NTC Business Books.
[39]E.H. Suh, K.C. Noh ,and C.K. Suh (1999), “Customer List Segmentation Using the Combined Response Model,” Expert Systems with Applications, Vol. 17, No. 2, pp. 89-97.
[40]Ho-Ha Sung and Sang-Chan Park(1998), “Application of Data Mining Tools to Hotel Data Mart on the Intranet for Database Marketing,” Expert Systems with Applications, Vol.15, pp. 1-31.
[41]Senecal Sylvain and Nantel Jacques (2004), “The Influence of Online Product Recommendations on Consumers’ Online Choices,” Journal of Retailing, Vol. 80, No. 2, pp. 159-169.
[42]A. Vellido, P.J.G. Lisboa ,and K. Meehan (1999), “Segmentation of the On-line Shopping Market Using Neural networks,” Expert Systems with Applications, Vol.17, pp. 303-314.
[43]Yi-Fan Wang, Yu-Liang Chuang, Mei-Hua Hsu ,and Huan-Chao Keh (2004), “A Personalized Recommender System for the Cosmetic Business,” Expert Systems with Applications , Vol. 26, No. 3, pp.427-434.
[44]Y. Wind (1978) , “Issues and Advances in Segmentation Research,” Journal of Marketing Research, Vol. 15, No.8, pp. 317-337.
[45] Soe-Tsyr Yuan and Y.W. Tsao (2003), “A Recommendation Mechanism for Contextualized Mobile Advertising,” Expert Systems with Applications, Vol. 24, No. 4, pp. 399-414.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔