請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/51659完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 盧信銘 | |
| dc.contributor.author | Yu-Ting Wang | en |
| dc.contributor.author | 王予廷 | zh_TW |
| dc.date.accessioned | 2021-06-15T13:43:18Z | - |
| dc.date.available | 2017-02-15 | |
| dc.date.copyright | 2016-02-15 | |
| dc.date.issued | 2015 | |
| dc.date.submitted | 2015-12-22 | |
| dc.identifier.citation | Adomavicius, G., & Tuzhilin, A. (2005). Toward the next generation of recommender systems: A survey of the state-of-the-art and possible extensions. Knowledge and Data Engineering, IEEE Transactions on, 17(6), 734-749.
Breese, J. S., Heckerman, D., & Kadie, C. (1998). Empirical analysis of predictive algorithms for collaborative filtering. Paper presented at the Proceedings of the Fourteenth conference on Uncertainty in artificial intelligence. Burke, R. (2002). Hybrid recommender systems: Survey and experiments. User Modeling and User-Adapted Interaction, 12(4), 331-370. Das, A. S., Datar, M., Garg, A., & Rajaram, S. (2007). Google news personalization: scalable online collaborative filtering. Paper presented at the Proceedings of the 16th International Conference on World Wide Web. Hofmann, T. (1999). Probabilistic latent semantic indexing. Paper presented at the Proceedings of the 22nd Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. Hofmann, T. (2004). Latent semantic models for collaborative filtering. ACM Transactions on Information Systems (TOIS), 22(1), 89-115. Hu, Y., Koren, Y., & Volinsky, C. (2008). Collaborative filtering for implicit feedback datasets. Paper presented at the Data Mining, 2008. ICDM'08. Eighth IEEE International Conference on. Koren, Y., Bell, R., & Volinsky, C. (2009). Matrix factorization techniques for recommender systems. Computer(8), 30-37. Li, Y., Hu, J., Zhai, C., & Chen, Y. (2010). Improving one-class collaborative filtering by incorporating rich user information. Paper presented at the Proceedings of the 19th ACM International Conference on Information and Knowledge Management. Pan, R., & Scholz, M. (2009). Mind the gaps: weighting the unknown in large-scale one-class collaborative filtering. Paper presented at the Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Pan, R., Zhou, Y., Cao, B., Liu, N. N., Lukose, R., Scholz, M., & Yang, Q. (2008). One-class collaborative filtering. Paper presented at the Data Mining, 2008. ICDM'08. Eighth IEEE International Conference on. Resnick, P., Iacovou, N., Suchak, M., Bergstrom, P., & Riedl, J. (1994). GroupLens: an open architecture for collaborative filtering of netnews. Paper presented at the Proceedings of the 1994 ACM Conference on Computer Supported Cooperative Work. Sarwar, B., Karypis, G., Konstan, J., & Riedl, J. (2001). Item-based collaborative filtering recommendation algorithms. Paper presented at the Proceedings of the 10th International Conference on World Wide Web. Srebro, N., & Jaakkola, T. (2003). Weighted low-rank approximations. Paper presented at the 20th International Conference on Machine Learning. Weiss, G. M. (2004). Mining with rarity: a unifying framework. ACM SIGKDD Explorations Newsletter, 6(1), 7-19. | |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/51659 | - |
| dc.description.abstract | 推薦系統廣泛應用於商業領域中,因為隨著資訊科技發展以及全球化,商品的可取用性已經不是使用者選擇商品的障礙了。然而使用者能夠實際使用的商品相較於其可接觸的商品卻十分有限。因此,而對於商品提供者而言,推薦系統的重要性在於,了解使用者購買的傾向,篩選出滿足使用者需求的商品。
推薦系統又以協同過濾的方法最常使用,協同過濾系統通常利用使用者對於產品的回饋,分析使用者對產品的喜好與彼此的關聯,藉此預測使用者對於某個項目的喜好程度進行推薦。但在現實狀況,我們經常只能取得使用者是否有採用該項目的紀錄,而不一定能蒐集到使用者對於該項目的喜好。另一方面,使用者對於項目的喜好也未必代表他就會採用某個產品。 本研究著重在只有一種回饋的單類別協同過濾問題中。在這類問題,只會以正向回饋來表示使用者是否採用了某個項目。這樣的資料具有一定的模糊程度,因為使用者未取用的項目包含了其所不想取用,以及其尚未接觸兩種可能。於是,我們透過加權的概念,去衡量未知回饋被認定負向回饋或未接觸的可信度。 本研究使用兩個大規模的電影資料庫預測使用者觀看的電影,並且實作間隙加權交替最小平方法 (Gap-weighting Alternating Least Square, gALS ),並調整加權設計,討論其對於模型的影響。 而透過我們的實驗結果可以發現交替最小平方法確實提供較好的預測能力,但是加權的策略卻沒有帶來巨大的改變,這或許說明目前採用權重設計太過簡單,無法代表所有的變數,因此建議未來能將成本敏感方法以及偏差性的概念納入權重設計當中。 | zh_TW |
| dc.description.abstract | Recommendation systems have been widely used in e-commerce applications. With the development of information technology, users can easily reach enormous products. However, users have limited ability to evaluate their choices. Therefore, it’s important for content providers and e-retailers to recommend items which meet users’ taste to enhance user satisfaction and loyalty. Collaborative filtering is a popular way to implement a recommendation system. Collaborative filtering analyzes the relationships between users and items by users’ feedback which reflect users’ preferences. Then, it recommends user a ranked item list which is sorted by predicted preferences.
This research focus on the One-class Collaborative Filtering (OCCF) approach. In OCCF, we only have positive examples to represent users’ actions. The data are ambiguous because unobserved data points can be interpreted as missing or negative cases. In this study, we treat unknown examples as negative examples with a confidence score, which is calculated by our weighting schemes. We apply our model on two large-scale movie rating datasets, and implement OCCF with gap-weighting Alternating Least Square (gALS). Then, we adjust weighting schemes to observe the impact on the model. Our result shows that gALS improves predicting performance. However, weighting strategies don’t make a dramatic impact. | en |
| dc.description.provenance | Made available in DSpace on 2021-06-15T13:43:18Z (GMT). No. of bitstreams: 1 ntu-104-R02725031-1.pdf: 3606648 bytes, checksum: a06dd0836a568534eb55f75444796482 (MD5) Previous issue date: 2015 | en |
| dc.description.tableofcontents | 目錄
口試委員會審定書 i 致謝 ii 中文摘要 iii Abstract iv 第一章 緒論 1 1.1 研究背景與動機 1 1.2 研究目的 3 第二章 文獻探討 4 2.1. 推薦系統 4 2.2 內容為基之推薦系統 5 2.3 協同過濾推薦系統 6 2.4 Memory-Based 協同過濾系統 8 2.5 Model-Based 協同過濾系統 9 2.5.1 隱因子模型 (Latent factor model) 9 2.5.2 交替最小平方法 (ALS) 11 2.5.3 奇異值分解 (Singular Value Decomposition) 11 2.5.4 機率潛在語意分析 (Probabilistic Latent Semantic Analysis, pLSA) 13 2.6 單類別協同過濾 (OCCF) 與隱性回饋 14 2.7 系統規模化 (Scalability) 17 第三章 系統設計 18 3.1 加權交替最小平方法 18 3.1.1 加權交替最小平方法 (wALS) 18 3.1.2 權重設計 19 3.1.3 ALS 與 wALS 的時間複雜度比較 20 3.1.4 間隙加權最小平方法 (Gap-weighting ALS, gALS) 22 3.2 基礎線模型 24 3.2.1 熱門度模型 24 3.2.2 低階近似SVD模型 24 3.2.3 pLSA模型 25 3.2.4 Item-based 協同過濾模型 25 第四章 研究資料概觀與資料處理 27 4.1 資料來源 27 4.1.1 Netflix Prize資料 27 4.1.2 MovieLens 資料 27 4.2 資料處理 28 第五章 實驗設計與結果 29 5.1衡量指標 29 5.2 5等分交叉驗證 (Five-Fold Cross Validation) 31 5.3 權重設計 32 5.4 實驗結果 38 5.5 收斂速度 40 5.6 以相反假設設計權重值 42 第六章 結論與建議 45 6.1 實驗結論 45 6.2 研究貢獻 46 6.3 未來研究方向 47 參考文獻 48 附錄A 50 | |
| dc.language.iso | zh-TW | |
| dc.subject | 協同過濾 | zh_TW |
| dc.subject | 矩陣分解 | zh_TW |
| dc.subject | 推薦系統 | zh_TW |
| dc.subject | 交替最小平方法 | zh_TW |
| dc.subject | 單類別協同過濾 | zh_TW |
| dc.subject | Netflix prize | en |
| dc.subject | matrix factorization | en |
| dc.subject | alternative least square | en |
| dc.subject | one-class collaborative filtering | en |
| dc.subject | collaborative filtering | en |
| dc.subject | Recommendation systems | en |
| dc.title | 應用於大量運算之單類別協同過濾:探討權重設計對推薦效能的影響 | zh_TW |
| dc.title | Large-Scale One-Class Collaborative Filtering:The Impact of Weighting Schemes | en |
| dc.type | Thesis | |
| dc.date.schoolyear | 104-1 | |
| dc.description.degree | 碩士 | |
| dc.contributor.oralexamcommittee | 魏志平,孔令傑 | |
| dc.subject.keyword | 協同過濾,推薦系統,單類別協同過濾,交替最小平方法,矩陣分解, | zh_TW |
| dc.subject.keyword | Recommendation systems,collaborative filtering,one-class collaborative filtering,alternative least square,matrix factorization,Netflix prize, | en |
| dc.relation.page | 54 | |
| dc.rights.note | 有償授權 | |
| dc.date.accepted | 2015-12-22 | |
| dc.contributor.author-college | 管理學院 | zh_TW |
| dc.contributor.author-dept | 資訊管理學研究所 | zh_TW |
| 顯示於系所單位: | 資訊管理學系 | |
文件中的檔案:
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| ntu-104-1.pdf 未授權公開取用 | 3.52 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。
