在推薦系統中，一般所用來訓練的評分資料相對整個分數矩陣所佔的比例非常小，導致推薦系統的訓練資料不足，無法將推薦效果彰顯，我們稱這樣的問題為資料過疏的問題。在過去，這些評分資料來自於顧客對產品主動評分，但在現實生活中，顧客未必會主動評分，因此而造成評分資料不足。我們希望有別於以往由顧客主動評分，採用顧客購買產品及瀏覽紀錄的資料當作訓練資料，這將會有效提升訓練資料量。在購買行為或瀏覽網頁的行為當中，我們只能得知使用者買了什麼或點選了什麼，而無法得到消費者不買什麼，因此這個新的資料矩陣為「一個類別（one-class）」的矩陣，這種一個類別的推薦問題，亦可以應用於網頁推薦、書籤推薦及社群網路之好友推薦等。傳統的評分矩陣，每個資料數值範圍為一到五分，而一個類別的資料矩陣則只存在正類別與負類別兩類，且訓練資料中只會有正類別資料。因此我們基於一般在分類問題上專門解決正範例與未分類範例之學習方法（Positive and Unlabeled-Learning）套用於一個類別的推薦系統當中（本文將稱我們所提供的方法為RPU）。由於在資料矩陣當中的資料並沒有正式的特徵以表示每筆資料，因此在本研究中，我們將針對資料矩陣找出有效的特徵，在我們所提出的四個特徵中，其中名為hybrid-MFsim的特徵選取方法在MovieLens及騰訊微博的資料都有不錯的表現，這些特徵選取方式也可以適用於一個類別的支持向量機分類模型（OCSVM）。我們比較了RPU與數個傳統針對一個類別的推薦方法，其中我們提出的RPU較其他方法能解決一個類別的推薦問題。

With the explosive growth of e-commerce, recommendation systems are getting well-known. Many kinds of recommendation systems are used in various websites. The one-class problem in recommendation systems is also a kind of recommendation problems that we cannot ignore. In the traditional one-class classification problem, we usually come up with Positive and Unlabeled Learning (PU-Learning). We cannot apply PU-Learning directly because we only can get the user-item rating matrix. In other words, the information is not enough to be applied to PU-Learning. The rating matrix does not have the available features. Without getting the outlier information of the users and the items, we have to define the features from the one-class rating matrix. In this paper, we proposed a PU-Learning framework which is focusing on the one-class recommendation problem, called RPU (Recommendation by PU-Learning). We use MovieLens and TencentWeibo datasets to evaluate our methods. RPU can get the best accuracy comparing with the baselines of one-class approaches. In RPU, we have a feature selection part. We also contribute a few kinds of features that are appropriate for RPU. The features not only perform great in our RPU framework but perform around 10% better by using One-Class SVM.

[1] M. Balabanović and Y. Shoham, "Fab: content-based, collaborative recommendation," Communications of the ACM, vol. 40, pp. 66-72, 1997.
[2] J. S. Breese, D. Heckerman, and C. Kadie, "Empirical analysis of predictive algorithms for collaborative filtering," presented at the Proceedings of the Fourteenth conference on Uncertainty in artificial intelligence, 1998.
[3] R. Burke, "Hybrid web recommender systems," in The adaptive web, ed: Springer, pp. 377-408,2007.
[4] C. C. Chang and C. J. Lin, "{LIBSVM}: A library for support vector machines," ACM Transactions on Intelligent Systems and Technology, vol. 2, pp. 27:1--27:27, 2011.
[5] Y. Chen, X. S. Zhou, and T. S. Huang, "One-class SVM for learning in image retrieval," in Image Processing, 2001. Proceedings. 2001 International Conference on, pp. 34-37, 2001.
[6] C. Cortes and V. Vapnik, "Support-vector networks," Machine learning, vol. 20, pp. 273-297, 1995.
[7] P. Cremonesi, Y. Koren, and R. Turrin, "Performance of recommender algorithms on top-n recommendation tasks," in Proceedings of the fourth ACM conference on Recommender systems, pp. 39-46, 2010.
[8] J. Delgado and N. Ishii, "Memory-based weighted majority prediction," in SIGIR Workshop Recomm. Syst. Citeseer, 1999.
[9] C. Elkan and K. Noto, "Learning classifiers from only positive and unlabeled data," in Proceedings of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining, pp. 213-220, 2008.
[10] W. Hill, L. Stead, M. Rosenstein, and G. Furnas, "Recommending and evaluating choices in a virtual community of use," in Proceedings of the SIGCHI conference on Human factors in computing systems, pp. 194-201, 1995.
[11] Y. Hu, Y. Koren, and C. Volinsky, "Collaborative filtering for implicit feedback datasets," in Data Mining, 2008. ICDM'08. Eighth IEEE International Conference on, pp. 263-272, 2008.
[12] D. Jannach, M. Zanker, A. Felfernig, and G. Friedrich, Recommender systems: an introduction: Cambridge University Press, 2010.
[13] J. A. Konstan, B. N. Miller, D. Maltz, J. L. Herlocker, L. R. Gordon, and J. Riedl, "GroupLens: applying collaborative filtering to Usenet news," Communications of the ACM, vol. 40, pp. 77-87, 1997.
[14] Y. Koren, R. Bell, and C. Volinsky, "Matrix factorization techniques for recommender systems," Computer, pp. 30-37, 2009.
[15] X.-L. Li, L. Zhang, B. Liu, and S.-K. Ng, "Distributional similarity vs. PU learning for entity set expansion," in Proceedings of the ACL 2010 Conference Short Papers, pp. 359-364, 2010.
[16] B. Liu, Web data mining: exploring hyperlinks, contents, and usage data: Springer Science & Business Media, 2007.
[17] B. Liu, W. S. Lee, P. S. Yu, and X. Li, "Partially supervised classification of text documents," in ICML, pp. 387-394, 2002.
[18] T. Mahmood and F. Ricci, "Improving recommender systems with adaptive conversational strategies," in Proceedings of the 20th ACM conference on Hypertext and hypermedia, pp. 73-82, 2009.
[19] L. M. Manevitz and M. Yousef, "One-class SVMs for document classification," the Journal of machine Learning research, vol. 2, pp. 139-154, 2002.
[20] M. M. Moya and D. R. Hush, "Network constraints and multi-objective optimization for one-class classification," Neural Networks, vol. 9, pp. 463-474, 1996.
[21] A. Nakamura and N. Abe, "Collaborative Filtering Using Weighted Majority Prediction Algorithms," in ICML, pp. 395-403, 1998.
[22] R. Pan and M. Scholz, "Mind the gaps: weighting the unknown in large-scale one-class collaborative filtering," in Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining, pp. 667-676, 2009.
[23] R. Pan, Y. Zhou, B. Cao, N. N. Liu, R. Lukose, M. Scholz, et al., "One-class collaborative filtering," in Data Mining, 2008. ICDM'08. Eighth IEEE International Conference on, pp. 502-511, 2008.
[24] M. J. Pazzani and D. Billsus, "Content-based recommendation systems," in The adaptive web, ed: Springer, pp. 325-341,2007.
[25] S. Rendle, C. Freudenthaler, Z. Gantner, and L. Schmidt-Thieme, "BPR: Bayesian personalized ranking from implicit feedback," in Proceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence, pp. 452-461, 2009.
[26] P. Resnick and H. R. Varian, "Recommender systems," Communications of the ACM, vol. 40, pp. 56-58, 1997.
[27] B. Sarwar, G. Karypis, J. Konstan, and J. Riedl, "Item-based collaborative filtering recommendation algorithms," in Proceedings of the 10th international conference on World Wide Web, pp. 285-295, 2001.
[28] V. Sindhwani, S. Bucak, J. Hu, and A. Mojsilovic, "A family of non-negative matrix factorizations for one-class collaborative filtering problems," in Proceedings of the ACM Recommender Systems Conference, RecSys, 2009.