臺灣博碩士論文加值系統

推薦系統根據使用者個人的偏好與需求，協助使用者從大量的選擇中，找到使用者感興趣或相關性高的資訊。以往的推薦系統大多傾向於使用單一準則評分的方法來進行資料分析，但經過此方法篩選而出的資訊並非能夠完全符合使用者的需求，這是由於評價物品的好壞並非只能用單一的準則來概括表示，為了改善此現象，本研究使用多準則評分的推薦系統來進行資料分析。在多準則評分的推薦系統中，使用者在一個項目中會有兩個或兩個以上的評分，而不同的使用者對於準則考量的評分權重也會有所不同。而以往多準則評分的推薦系統決定彙總函數的技術大多是採用領域專長，但是如此決定的彙總函數太過主觀。因此本研究使用統計技術中的線性迴歸來決定使用者的個別準則權重，使產生的推薦結果能夠更加符合使用者的需求，並提高推薦系統的準確率。本研究使用準確率、喚回率與F1-measure準則對模型進行評估。其實驗結果顯示本研究提出整合準則權重於多準則協同過濾推薦模型顯著的優於傳統協同過濾技術與均一權重技術。

Recommendation system according to user individual preferences and needs, as-sists the user from massive choices, found the user to be interested or the high related information.Traditional recommendation systems often use single criteria rating ap-proach for data analysis, but after filtering out the information in this method is not able to meet the needs of users, this is because the rating item of good or bad is not only use a single criteria rating that, in order to improve this situation, the study recommended using Multi-criteria ratings recommendation system for data analysis. In a multi-criteria ratings recommendation system, the user in a item there will be two or more ratings, while users of different criteria for consideration for the weight ratings will be different. The previous multi-criteria ratings recommendation system by most of the decision aggregate function is used in the field of domain expertise, but such decisions are too subjective aggregate function. This paper used statistical tech-niques in the linear regression to determine the weight of the user's individual criteria, so that can produce results more in line with the recommendation of the user's needs and to improve the accuracy of recommendation systems. In this paper, the precision, recall and F1-measure criteria for assess of the model. The experimental results show that the proposed integrating weights into multi-criteria collaborative filtering recommendation model significantly better than the traditional collaborative filtering technology and same weight technology.

中文摘要 ．．．．．．．．．．．．．．．．．．．．． iii
英文摘要　 ．．．．．．．．．．．．．．．．．．．． iv
誌謝辭　　 ．．．．．．．．．．．．．．．．．．．． vi
內容目錄　 ．．．．．．．．．．．．．．．．．．．． vii
表目錄　　 ．．．．．．．．．．．．．．．．．．．． ix
圖目錄　　 ．．．．．．．．．．．．．．．．．．．． x
第一章　　緒論 ．．．．．．．．．．．．．．．．．． 1
　　第一節　　研究背景 ．．．．．．．．．．．．．． 1
　　第二節　　研究動機 ．．．．．．．．．．．．．． 2
　　第三節　　研究目的 ．．．．．．．．．．．．．． 3
　　第四節　　研究範圍與限制 ．．．．．．．．．．． 4
　　第五節　　論文架構與研究流程 ．．．．．．．．． 5
第二章　　文獻探討　 ．．．．．．．．．．．．．．． 7
　　第一節　　內容過濾推薦 ．．．．．．．．．．．． 7
　　第二節　　協同過濾推薦 ．．．．．．．．．．．． 10
　　第三節　　混合式推薦 ．．．．．．．．．．．．． 16
　　第四節　　多準則決策 ．．．．．．．．．．．．． 17
　　第五節　　迴歸分析 ．．．．．．．．．．．．．． 23
　　第六節　　多準則評分推薦 ．．．．．．．．．．． 25
第三章　　研究方法　 ．．．．．．．．．．．．．．． 29
　　第一節　　系統架構 ．．．．．．．．．．．．．． 29
　　第二節　　多準則權重計算模組 ．．．．．．．．． 32
　　第三節　　推薦模組 ．．．．．．．．．．．．．． 36
第四章　　實驗與結果分析 ．．．．．．．．．．．．． 38
　　第一節　　實驗環境與工具 ．．．．．．．．．．． 38
　　第二節　　評估方法 ．．．．．．．．．．．．．． 38
　　第三節　　電影研究設計 ．．．．．．．．．．．． 39
　　第四節　　實驗結果 ．．．．．．．．．．．．．． 40
第五章　　結論與未來展望 ．．．．．．．．．．．．． 51
　　第一節　　結論與貢獻 ．．．．．．．．．．．．． 51
　　第二節　　未來展望 ．．．．．．．．．．．．．． 52
參考文獻　 ．．．．．．．．．．．．．．．．．．．． 53

一、中文部分

周子敬，黃廷合(2007)，統計套裝軟體－精通SPSS，台北：全華科技圖書股份有限公司。

陳亭光(2008)，基於使用者經驗之多準則評分遊戲推薦系統，台灣大學資訊管理研究所未出版之碩士論文。

二、英文部分

Adomavicius, G., & Kwon, Y. (2007). New recommendation tech-niques for multicriteria rating systems. IEEE Intelligent Sys-tems, 22(3), 48-55.

Adomavicius, G., & Tuzhilin, A. (2005). Toward the next generation of recommender systems: A survey of the state-of-the-art and possible extensions. IEEE Transactions on Knowledge and Data Engineering, 17(6), 734-749.

Balabnovic, B., & Shoham, Y. (1997). Fab: Content-based, collabo-rative recommendation. Communications of the ACM, 40(3), 66-72.

Breese, J. S., Heckerman, D., & Kadie, C. (1998). Empirical analysis of predictive algorithms for collaborative filtering. In P. Ndilikilikesha (Ed.), Proceedings of the 14th Conference on Uncertainty in Artificial Intelligence (pp. 43-52). San Franciso: Morgan Kaufmann.

Goldberg, D., Nichols, D., Oki, B. M., & Terry, D. (1992). Using collaborative filtering to weave an information tapestry. Com-munications of the ACM, 35(12), 61-70.

Hwang, C. L., Lai, Y. J., & Liu, T. Y. (1993). A new approach for multiple objectivedecision making. Computers Operation Re-search, 20(8), 889-899.

Hwang, C. L., & Yoon, K. (Eds.). (1981). Multiple attribute decision making: Methods and applications. Heidelberg, Berlin: Springer-Verlag.

Koopmans, T. C. (1951). Activity analysis of production and alloca-tion. New York: John Wiley.

Kuhn, H. W., & Tucker, A. W. (1591). Nonlinear programming. In J. Neyman (Ed.), Proceedings of the Second Berkeley Symposium on mathematical Statistics and Probability (pp. 481-492). Berkeley, Calif: University of California Press.

Lakiotaki, K., Tsafarakis, S., & Matsatsinis, N. (2008). UTA-Rec: A recommender system based on multiple criteria analysis. In P. Pu, D. G. Bridge, B. Mobasher, F. Ricci (Eds.), Proceedings of the 2008 ACM conference on Recommender systems (pp. 219-226). Lausanne, Switzerland: ACM Press.

Manouselis, N., & Costopoulou, C. (2007). Analysis and classifica-tion of multi-criteria recommender systems. World Wide Web, 10(4), 415-441.

Nnadi, N. (2009). Applying relevant set correlation clustering to multi-criteria recommender systems. In L. D. Bergman, A. Tuzhilin, R. D. Burke, A. Felfernig, L. S. Thieme (Eds.), Pro-ceedings of the third ACM conference on Recommender sys-tems (pp. 401-404). New York: ACM Press.

Pazzani, M., & Billsus, D. (1997). Learning and revising user pro-files: The identification of interesting web sites. Machine Learning, 27(3), 313-331.

Resnick, P., Iacovou, N., Suchak, M., Bergstrom, P., & Riedl, J. (1994). GroupLens: An open architecture for collaborative fil-tering of netnews. In the Association for Computing Machin-ery (Ed.), Proceeding of ACM 1994 Conference of Computer Supported Cooperative Work (pp. 175-186). San Francisco: Chapel Hill.

Resnick, P., & Varian, H. R. (1997). Recommender systems. Com-munications of the ACM, 40(3), 56-58.

Roy, B. (1991). The outranking approach and the foundations of elec-tre methods. Theory and Decision, 31(1), 49-73.

Roy, B., & Vincke, P. (1981). Multicriteria analysis: Survey and direc-tion. European Journal of Operational Research, 8(3), 207-218.

Saaty, T. L. (1980). The analytic hierarchy process. New York: McGraw Hill.

Sarwar, B., Karypis, G., Konstan, J., & Riedl, J. (2000). Analysis of recommender algorithms for e-commerce. In the Association for Computing Machinery (Ed.), Proceedings of the Second ACM E-commerce Conference (p. 161), Minneapolis: IBM Institute for Advanced Commerce.

Salton, G., & Buckley, C. (1988). Term-weighting approaches in automatic text retrieval. Cornell University Technical Report, 24(5), 513-523.

Shardanand, U., & Maes, P. (1995). Social information filtering: Al-gorithms for automating ‘word of mouth’. In Com-puter-Human Interaction (Ed.), Proceedings of the Conference on Human Factors in Computing Systems (pp. 210-217). New York: ACM Press.