學術產出-學位論文
文章檢視/開啟
書目匯出
-
題名 網路評比資料之統計分析
Statistical analysis of online rating data作者 張孫浩 貢獻者 翁久幸
Weng, Chui Hsing
張孫浩關鍵詞 線上評分
推薦系統
IRT模型法
相關係數法
矩陣分解
online rating
recommender system
IRT model-based method
method, correlation-coefficient method
matrix factorization日期 2010 上傳時間 5-九月-2013 15:12:30 (UTC+8) 摘要 隨著網路的發達,各式各樣的資訊和商品也在網路上充斥著,使用者尋找資訊或是上網購物時,有的網站有推薦系統(recommender system)能提供使用者相關資訊或商品。若推薦系統能夠讓消費者所搜尋的相關資訊或商品能夠符合他們的習性時,便能讓消費者增加對系統的信賴程度,因此系統是否能準確預測出使用者的偏好就成為一個重要的課題。本研究使用兩筆資料,並以相關研究的三篇文獻進行分析和比較。這三篇文獻分別為IRT模型法(IRT model-based method)、相關係數法(correlation-coefficient method)、以及矩陣分解法(matrix factorization)。在經過一連串的實證分析後,歸納出以下結論:1. 模型法在預測方面雖然精確度不如其他兩種方法來的好,但是模型有解釋變數之間的關係以及預測機率的圖表展示,因此這個方法仍有存在的價值。2. 相關係數法容易因為評分稀疏性的問題而無法預測,建議可以搭配內容式推薦系統的運作方式協助推薦。3. 矩陣分解法在預測上雖然比IRT模型法還好，但分量的數字只是一個最佳化的結果，實際上無法解釋這些分量和數字的意義。
With the growth of the internet, websites are full of a variety of information and products. When users find the information or surf the internet to shopping, some websites provide users recommender system to find with which related. Hence, whether the recommender system can predict the users` preference is an important topic. This study used two data,which are "Mondo" and "MovieLens", and we used three related references to analyze and compare them. The three references are following: IRT model-based method, Correlation-coefficient method, and Matrix factorization.After the data analysis, we get the following conclusions:1. IRT model-based method is worse then other methods in predicting, but it can explain the relationship of variables and display the graph of predicting probabilities. Hence this method still has it`s value.2. Correlation-coefficient method is hard to predict because of sparsity. We can connect it with content filtering approach.3. Although matrix factorization is better then IRT model-based method in predicting, the vectors is a result of optimization. It may be hard to explain the meaning of the vectors.參考文獻 Agresti, A. ,"An Introduction to Catogerical Data Analysis," Wiley-Introduction.Cheung, K. , Tsui K. ,and Liu J. (2004), "Extended Latent Class Models for Collaborative Recommendation," IEEE Transactions on Systems, Man & Cybernetics: Part A, Jan 2004, Vol.34, Issue 1, pp. 143-148.Conry, D. C. (2009), "Recommender Systems for the Conference Paper Assignment Problem," thesis submitted to the Faculty of the Virginia Polytechnic Institute and State University.Ho, D. E. ,and Quinn, K. M. (2008), "Improving the Presentation and Interpretation of Online Ratings Data with Model-Based Figures," The American Statistician, Nov 2008, Vol.62, Issue 4, pp. 279-288.Kagie, M. ,Loos, M. ,and Wezel, M. (2009), "Including item characteristics in the probabilistic latent semantic analysis model for collaborative filtering," AI Communications, 22, 2009, pp. 249-265.Konstan, J. A. ,Miller, B. N. ,Maltz, D. ,Herlocker, J. L. ,Gordon, L. R.,and Riedl, J. (1997), "GroupLens: Applying Collaborative Filtering to Usenet News," Comminications of the ACM, Mar1997, Vol.40, Issue 3, pp. 77-87.Koren, K. ,Bell, R. ,and Volinsky, C. (2009), "Matrix Factorization Techniques for Recommender Systems," IEEE Computer Society, Aug 2009, Vol.42, Issue 8, pp. 42-49. Koren, K. (2010), "Collaborative Filtering with Temporal Dynamics," Comminications of the ACM, Apr 2010, Vol.53, Issue 4, pp. 89-98.Li, W. ,Lee, K. ,and Leung, K. (2006), "Generalized Regularized Least-Squares Learning with Predefined Features in a Hilbert Space," Neural Information Processing Systems - NIPS, pp. 881-888Resnick, P. , Iacovou, N. ,Suchak, M. ,Bergstrom, P. and Riedl, J. (1994), "GroupLens: An Open Architecture for Collaborative Filtering of Netnews," Proceedings of ACM 1994 Conference on Computer Supported Cooperative Work, Chapel Hill, pp. 175-186.Weisberg, S. "Applied linear regression," Wiley-Introduction.Williamson, S. and Ghahramani, Z. (2008),"Probabilistic Models for Data Combination in Recommender Systems," Probabilistic models for data combination in recommender systems In: Learning from Multiple Sources Workshop, 8-12 December 2008, Vancouver and Whistler, British Columbia, Canada.Zhou, H. ,and Lange, K. (2009), "Rating Movies and Rating the Raters Who Rate Them," The American Statistician, pp. 297-307.馮文正 (2001),合作式網站推薦系統,國立交通大學資訊科學所碩士論文吳肇銘 (2004),以消費者購買決策為基礎之適性化推薦系統,中原大學資訊管理學系碩士論文Amazon. Retrieved Nov, 2010, from http://www.amazon.comMondo Times. Retrieved Nov, 2010, from http:///www.mondotimes.comNetflix. Retrieved Nov, 2010, from http:///www.netflix.comPC magazine. Retrieved Nov, 2010, from http://www.pcmag.comTiVo台灣網站. Retrieved Nov, 2010, from http://www.tgc-taiwan.com.tw/index.php 描述 碩士
國立政治大學
統計研究所
98354010
99資料來源 http://thesis.lib.nccu.edu.tw/record/#G0098354010 資料類型 thesis dc.contributor.advisor 翁久幸 zh_TW dc.contributor.advisor Weng, Chui Hsing en_US dc.contributor.author (作者) 張孫浩 zh_TW dc.creator (作者) 張孫浩 zh_TW dc.date (日期) 2010 en_US dc.date.accessioned 5-九月-2013 15:12:30 (UTC+8) - dc.date.available 5-九月-2013 15:12:30 (UTC+8) - dc.date.issued (上傳時間) 5-九月-2013 15:12:30 (UTC+8) - dc.identifier (其他 識別碼) G0098354010 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/60440 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 統計研究所 zh_TW dc.description (描述) 98354010 zh_TW dc.description (描述) 99 zh_TW dc.description.abstract (摘要) 隨著網路的發達,各式各樣的資訊和商品也在網路上充斥著,使用者尋找資訊或是上網購物時,有的網站有推薦系統(recommender system)能提供使用者相關資訊或商品。若推薦系統能夠讓消費者所搜尋的相關資訊或商品能夠符合他們的習性時,便能讓消費者增加對系統的信賴程度,因此系統是否能準確預測出使用者的偏好就成為一個重要的課題。本研究使用兩筆資料,並以相關研究的三篇文獻進行分析和比較。這三篇文獻分別為IRT模型法(IRT model-based method)、相關係數法(correlation-coefficient method)、以及矩陣分解法(matrix factorization)。在經過一連串的實證分析後,歸納出以下結論:1. 模型法在預測方面雖然精確度不如其他兩種方法來的好,但是模型有解釋變數之間的關係以及預測機率的圖表展示,因此這個方法仍有存在的價值。2. 相關係數法容易因為評分稀疏性的問題而無法預測,建議可以搭配內容式推薦系統的運作方式協助推薦。3. 矩陣分解法在預測上雖然比IRT模型法還好，但分量的數字只是一個最佳化的結果，實際上無法解釋這些分量和數字的意義。 zh_TW dc.description.abstract (摘要) With the growth of the internet, websites are full of a variety of information and products. When users find the information or surf the internet to shopping, some websites provide users recommender system to find with which related. Hence, whether the recommender system can predict the users` preference is an important topic. This study used two data,which are "Mondo" and "MovieLens", and we used three related references to analyze and compare them. The three references are following: IRT model-based method, Correlation-coefficient method, and Matrix factorization.After the data analysis, we get the following conclusions:1. IRT model-based method is worse then other methods in predicting, but it can explain the relationship of variables and display the graph of predicting probabilities. Hence this method still has it`s value.2. Correlation-coefficient method is hard to predict because of sparsity. We can connect it with content filtering approach.3. Although matrix factorization is better then IRT model-based method in predicting, the vectors is a result of optimization. It may be hard to explain the meaning of the vectors. en_US dc.description.tableofcontents 第一章 緒論 page 61.1節 研究背景 page 61.1.1節 推薦系統簡介 page 71.1.2節 目前網路評分呈現的瑕疵 page 81.2節 研究目的 page 9第二章 文獻回顧 page 102.1節 IRT模型法 page 102.2節 相關係數預測法 page 132.3節 矩陣分解理論 page 15第三章 研究方法 page 183.1節 IRT模型法 page 183.2節 相關係數預測法 page 193.3節 矩陣分解法 page 20第四章 實證研究 page 214.1節 實證資料 page 214.2節 IRT模型法分析 page 224.2.1節 beta_r是否要大於0 page 224.2.2節 Mondo預測結果 page 254.2.3節 MovieLens預測結果 page 334.2.4節 IRT模型法的改進 page 374.3節 相關係數法分析 page 384.3.1節 Mondo預測結果 page 384.3.2節 MovieLens預測結果 page 404.4節 矩陣分解法分析 page 424.4.1節 Mondo預測結果 page 424.4.2節 MovieLens預測結果 page 424.5節 預測結果比較 page 45第五章 結論與建議 page 47 zh_TW dc.format.extent 1462105 bytes - dc.format.mimetype application/pdf - dc.language.iso en_US - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0098354010 en_US dc.subject (關鍵詞) 線上評分 zh_TW dc.subject (關鍵詞) 推薦系統 zh_TW dc.subject (關鍵詞) IRT模型法 zh_TW dc.subject (關鍵詞) 相關係數法 zh_TW dc.subject (關鍵詞) 矩陣分解 zh_TW dc.subject (關鍵詞) online rating en_US dc.subject (關鍵詞) recommender system en_US dc.subject (關鍵詞) IRT model-based method en_US dc.subject (關鍵詞) method, correlation-coefficient method en_US dc.subject (關鍵詞) matrix factorization en_US dc.title (題名) 網路評比資料之統計分析 zh_TW dc.title (題名) Statistical analysis of online rating data en_US dc.type (資料類型) thesis en dc.relation.reference (參考文獻) Agresti, A. ,"An Introduction to Catogerical Data Analysis," Wiley-Introduction.Cheung, K. , Tsui K. ,and Liu J. (2004), "Extended Latent Class Models for Collaborative Recommendation," IEEE Transactions on Systems, Man & Cybernetics: Part A, Jan 2004, Vol.34, Issue 1, pp. 143-148.Conry, D. C. (2009), "Recommender Systems for the Conference Paper Assignment Problem," thesis submitted to the Faculty of the Virginia Polytechnic Institute and State University.Ho, D. E. ,and Quinn, K. M. (2008), "Improving the Presentation and Interpretation of Online Ratings Data with Model-Based Figures," The American Statistician, Nov 2008, Vol.62, Issue 4, pp. 279-288.Kagie, M. ,Loos, M. ,and Wezel, M. (2009), "Including item characteristics in the probabilistic latent semantic analysis model for collaborative filtering," AI Communications, 22, 2009, pp. 249-265.Konstan, J. A. ,Miller, B. N. ,Maltz, D. ,Herlocker, J. L. ,Gordon, L. R.,and Riedl, J. (1997), "GroupLens: Applying Collaborative Filtering to Usenet News," Comminications of the ACM, Mar1997, Vol.40, Issue 3, pp. 77-87.Koren, K. ,Bell, R. ,and Volinsky, C. (2009), "Matrix Factorization Techniques for Recommender Systems," IEEE Computer Society, Aug 2009, Vol.42, Issue 8, pp. 42-49. Koren, K. (2010), "Collaborative Filtering with Temporal Dynamics," Comminications of the ACM, Apr 2010, Vol.53, Issue 4, pp. 89-98.Li, W. ,Lee, K. ,and Leung, K. (2006), "Generalized Regularized Least-Squares Learning with Predefined Features in a Hilbert Space," Neural Information Processing Systems - NIPS, pp. 881-888Resnick, P. , Iacovou, N. ,Suchak, M. ,Bergstrom, P. and Riedl, J. (1994), "GroupLens: An Open Architecture for Collaborative Filtering of Netnews," Proceedings of ACM 1994 Conference on Computer Supported Cooperative Work, Chapel Hill, pp. 175-186.Weisberg, S. "Applied linear regression," Wiley-Introduction.Williamson, S. and Ghahramani, Z. (2008),"Probabilistic Models for Data Combination in Recommender Systems," Probabilistic models for data combination in recommender systems In: Learning from Multiple Sources Workshop, 8-12 December 2008, Vancouver and Whistler, British Columbia, Canada.Zhou, H. ,and Lange, K. (2009), "Rating Movies and Rating the Raters Who Rate Them," The American Statistician, pp. 297-307.馮文正 (2001),合作式網站推薦系統,國立交通大學資訊科學所碩士論文吳肇銘 (2004),以消費者購買決策為基礎之適性化推薦系統,中原大學資訊管理學系碩士論文Amazon. Retrieved Nov, 2010, from http://www.amazon.comMondo Times. Retrieved Nov, 2010, from http:///www.mondotimes.comNetflix. Retrieved Nov, 2010, from http:///www.netflix.comPC magazine. Retrieved Nov, 2010, from http://www.pcmag.comTiVo台灣網站. Retrieved Nov, 2010, from http://www.tgc-taiwan.com.tw/index.php zh_TW
