基于半监督学习的个性化推荐研究
|
摘要
随着社交网络和电子商务等互联网技术的发展，人们逐渐从信息匮乏的时代步入“信息超载”的时代。海量信息在给用户带来极大便利的同时，也使用户迷失在信息的海洋中，很难找到自己感兴趣的信息。个性化推荐是解决该问题最有效的工具，它通过主动挖掘用户的兴趣偏好，为用户推送个性化的信息。
当前，主流的个性化推荐方法包括：基于协同过滤的方法和基于内容的方法。协同过滤的方法通过计算用户兴趣偏好的相似性，从而为目标用户过滤和筛选感兴趣的物品，它主要是基于用户的行为信息进行推荐，而没有真正利用物品的内容信息和用户的标签信息，同时也存在着数据稀疏和冷启动等问题；基于内容的推荐本质上则是一种信息过滤技术，仅仅通过学习用户历史选择的物品信息，缺乏对用户反馈信息的挖掘，这也往往会造成推荐结果过度特殊化。
针对上述推荐方法存在的问题，本文提出了利用半监督学习的方法实现基于用户行为信息与物品内容信息的个性化推荐。其主要工作如下：
①针对协同过滤推荐方法存在计算相似度方式单一等问题，提出了基于距离度量与高斯混合模型的半监督聚类的推荐方法。传统的协同过滤方法时间复杂度和用户数的增长近似于平方关系，当用户数很大时，计算非常耗时。本文提出利用聚类分析的方法替代用户兴趣的相似度计算，且综合考虑了用户行为偏好和物品内容信息。具体在聚类分析中，算法不仅考虑了数据的几何特征，也兼顾了数据的正态分布信息。
②针对个性化推荐中用户兴趣标签偏少的问题，提出了基于主动学习和协同训练的半监督推荐方法。传统的基于分类模型的推荐方法，当有标签数据偏少时，对挖掘用户潜在兴趣偏好非常不利，本文利用主动学习的策略抽取数据集中具有最大信息量的样本，通过咨询（Query）方式或领域专家标注的方式获得相应的标签，增加了训练模型的样本空间，以改进个性化推荐的质量。
③针对主动学习的方法加重了用户的负担或增加了人力成本的问题，提出了基于高斯对称分布的自增量学习的半监督推荐方法。该方法充分利用了大量的无标签的数据，并结合一定的有标签数据进行建模。具体在算法中，通过挑选具有高置信度且高斯对称分布的数据进行自增量学习，以改进个性化推荐的质量。
④针对在构建特征向量过程中，用户行为特征与物品内容特征的权重不易权衡的问题，提出了基于图模型的半监督推荐方法。算法通过SELF等方法计算权衡因子，且根据用户的行为信息构造基于最近邻图的权重矩阵。算法利用Sigmoid映射函数来度量两个用户的兴趣相似度，并在算法的损失函数中包括用户行为相似性约束和物品内容相似性约束，且两部分约束的权重由一个平衡因子权衡。
With the development of social networking, e-commerce and other Internettechnology, people gradually got into the “information overload” era from lack ofinformation era. It brings great convenience to users with vast amounts of information,but also let the user to get lost in ocean of information, it is difficult to find theinformation that they are interested in. Personalized recommended is used to pushpersonalized information for usersby mining user preferences, which is also as the mosteffective tool to solve the problem of information overload.
Currently, themainstream personalized recommendation methodsinclude:collaborative filtering method and content-based method. The key of collaborativefiltering method is the similarity computing of user interest preference,which is to filterinterest items for the target user. It makes recommendationsbased on the user’s behaviorinformationmainly, but did not take advantage of the item content information and userlabel information really. Meanwhile, there aredata sparse and cold start problems.Content-basedrecommendation is an information filtering technologyin essence, whichusers only learn the history of selected items of informationsimply.It can not mining theuser feedback information on items, which often leads to excessive specialization of therecommendation result.
According to the problem of above recommendation methods, the semi-supervisedlearning methods are proposedto achieve personalized recommendation based on userbehavior information and items content information. Thedetails of research work asfollows:
①According to these problemsthat traditional collaborative filtering algorithm issingle in the way of calculating the similarity,the semi-supervised hybrid clusteringbased on the distance metric and Gaussian model is proposed to slove these problems.The time complexity of the traditional collaborative filtering algorithm isquadraticof thenumber of users, when the number of users is large, it is time-consuming.In this paper,the cluster analysis is used to alternative the similarity computingof user interest, whichconsiders the preferences of the user behavior and content information. Specific tocluster analysis, the algorithm takes into account not only the geometric information ofdata samples, but also take into account the normal distributioninformation of datasamples.
②According to theproblem thatthe labeled dataof user interestis too few inpersonalized recommendation, the semi-supervisedrecommended method based onactive learning and collaborative training is proposed.About thetraditionalrecommended method based onclassification model, it has very negativepotential interest problem on mining user preferences whenthe labeled datais few. In thepaper, the user behavior information and item content information is used to model,andthe unlabeled data with largest amount of information is extracted with theactivelearning strategies, which increase the sample space of training modelby query mode orlabel the unlabeled data by field experts, to improve the quality of personalizedrecommendation.
③According to theproblem of it increase the burden on user or labor cost with theactive learning method, a personalized recommendation method withsemi-supervisedincremental learningbased on Gaussian symmetrical distribution isproposed.We use the large number data of no user tag information, combined with asmall amount of user tag data to build model. In the algorithm, the data selectionalgorithm chooses the unlabeled data with high confidenceand Gaussian symmetricaldistribution to iterative learning, to improve the quality of personalizedrecommendation.
④According to the problem that it is difficult to measure the features vectorweights between user behavior information and item content information,thesemi-supervisedrecommended method of graph-basedis proposed, which cancalculate the weighing factors with SELF method and other methods. The algorithmconstruct the weight matrix based on nearest neighbor graph with user behaviorinformation. Specially, Sigmoid mapping function is used to measure the interest degreeof two users; wedefine the loss function of the algorithm that includes user behaviorsimilarity constraints and item content similarity constraints, and the constraints of thesetwo parts are weightedwith a balance factor.
当前，主流的个性化推荐方法包括：基于协同过滤的方法和基于内容的方法。协同过滤的方法通过计算用户兴趣偏好的相似性，从而为目标用户过滤和筛选感兴趣的物品，它主要是基于用户的行为信息进行推荐，而没有真正利用物品的内容信息和用户的标签信息，同时也存在着数据稀疏和冷启动等问题；基于内容的推荐本质上则是一种信息过滤技术，仅仅通过学习用户历史选择的物品信息，缺乏对用户反馈信息的挖掘，这也往往会造成推荐结果过度特殊化。
针对上述推荐方法存在的问题，本文提出了利用半监督学习的方法实现基于用户行为信息与物品内容信息的个性化推荐。其主要工作如下：
①针对协同过滤推荐方法存在计算相似度方式单一等问题，提出了基于距离度量与高斯混合模型的半监督聚类的推荐方法。传统的协同过滤方法时间复杂度和用户数的增长近似于平方关系，当用户数很大时，计算非常耗时。本文提出利用聚类分析的方法替代用户兴趣的相似度计算，且综合考虑了用户行为偏好和物品内容信息。具体在聚类分析中，算法不仅考虑了数据的几何特征，也兼顾了数据的正态分布信息。
②针对个性化推荐中用户兴趣标签偏少的问题，提出了基于主动学习和协同训练的半监督推荐方法。传统的基于分类模型的推荐方法，当有标签数据偏少时，对挖掘用户潜在兴趣偏好非常不利，本文利用主动学习的策略抽取数据集中具有最大信息量的样本，通过咨询（Query）方式或领域专家标注的方式获得相应的标签，增加了训练模型的样本空间，以改进个性化推荐的质量。
③针对主动学习的方法加重了用户的负担或增加了人力成本的问题，提出了基于高斯对称分布的自增量学习的半监督推荐方法。该方法充分利用了大量的无标签的数据，并结合一定的有标签数据进行建模。具体在算法中，通过挑选具有高置信度且高斯对称分布的数据进行自增量学习，以改进个性化推荐的质量。
④针对在构建特征向量过程中，用户行为特征与物品内容特征的权重不易权衡的问题，提出了基于图模型的半监督推荐方法。算法通过SELF等方法计算权衡因子，且根据用户的行为信息构造基于最近邻图的权重矩阵。算法利用Sigmoid映射函数来度量两个用户的兴趣相似度，并在算法的损失函数中包括用户行为相似性约束和物品内容相似性约束，且两部分约束的权重由一个平衡因子权衡。
With the development of social networking, e-commerce and other Internettechnology, people gradually got into the “information overload” era from lack ofinformation era. It brings great convenience to users with vast amounts of information,but also let the user to get lost in ocean of information, it is difficult to find theinformation that they are interested in. Personalized recommended is used to pushpersonalized information for usersby mining user preferences, which is also as the mosteffective tool to solve the problem of information overload.
Currently, themainstream personalized recommendation methodsinclude:collaborative filtering method and content-based method. The key of collaborativefiltering method is the similarity computing of user interest preference,which is to filterinterest items for the target user. It makes recommendationsbased on the user’s behaviorinformationmainly, but did not take advantage of the item content information and userlabel information really. Meanwhile, there aredata sparse and cold start problems.Content-basedrecommendation is an information filtering technologyin essence, whichusers only learn the history of selected items of informationsimply.It can not mining theuser feedback information on items, which often leads to excessive specialization of therecommendation result.
According to the problem of above recommendation methods, the semi-supervisedlearning methods are proposedto achieve personalized recommendation based on userbehavior information and items content information. Thedetails of research work asfollows:
①According to these problemsthat traditional collaborative filtering algorithm issingle in the way of calculating the similarity,the semi-supervised hybrid clusteringbased on the distance metric and Gaussian model is proposed to slove these problems.The time complexity of the traditional collaborative filtering algorithm isquadraticof thenumber of users, when the number of users is large, it is time-consuming.In this paper,the cluster analysis is used to alternative the similarity computingof user interest, whichconsiders the preferences of the user behavior and content information. Specific tocluster analysis, the algorithm takes into account not only the geometric information ofdata samples, but also take into account the normal distributioninformation of datasamples.
②According to theproblem thatthe labeled dataof user interestis too few inpersonalized recommendation, the semi-supervisedrecommended method based onactive learning and collaborative training is proposed.About thetraditionalrecommended method based onclassification model, it has very negativepotential interest problem on mining user preferences whenthe labeled datais few. In thepaper, the user behavior information and item content information is used to model,andthe unlabeled data with largest amount of information is extracted with theactivelearning strategies, which increase the sample space of training modelby query mode orlabel the unlabeled data by field experts, to improve the quality of personalizedrecommendation.
③According to theproblem of it increase the burden on user or labor cost with theactive learning method, a personalized recommendation method withsemi-supervisedincremental learningbased on Gaussian symmetrical distribution isproposed.We use the large number data of no user tag information, combined with asmall amount of user tag data to build model. In the algorithm, the data selectionalgorithm chooses the unlabeled data with high confidenceand Gaussian symmetricaldistribution to iterative learning, to improve the quality of personalizedrecommendation.
④According to the problem that it is difficult to measure the features vectorweights between user behavior information and item content information,thesemi-supervisedrecommended method of graph-basedis proposed, which cancalculate the weighing factors with SELF method and other methods. The algorithmconstruct the weight matrix based on nearest neighbor graph with user behaviorinformation. Specially, Sigmoid mapping function is used to measure the interest degreeof two users; wedefine the loss function of the algorithm that includes user behaviorsimilarity constraints and item content similarity constraints, and the constraints of thesetwo parts are weightedwith a balance factor.
引文
[1] Information Overload[EB/OL]. http://en.wikipedia.org/wiki/Information_overload.
[2] Borkar V, Carey M J, Li C. Inside “Big Data management”: Ogres, Onions, or Parfaits?[C]. InProceedings of the15thInternational Conference on Extending Database Technology,2012:3-14.
[3]项亮编,陈义,王益审校.推荐系统实践[M].北京:人民邮电出版社,2013.
[4] Adomavicius G, Tuzhilin A. Towards the Next Generation of Recommender Systems: A Surveyof the State-of-the-Art and Possible Extensions [J]. IEEE Transactions on Knowledge and DataEngineering,2005,17(6):734–749.
[5]吴金龙. Netflix Prize中的协同过滤算法[D].北京:北京大学,2010.
[6] Sun J T, Zeng H J, Liu H, et al. CubeSVD: a Novel Approach to Personalized Web Search[C].In Proceedings of the14th International Conference on World Wide Web,2005:382-390.
[7] Sun J T, Wang X H, Shen D, et al. Mining Click through Data for Collaborative Web Search [C].In Proceedings ofthe15th International Conference on World Wide Web,2006:947-948.
[8] Rich E. User Modeling via Stereotypes [J]. Cognitive Science,1979,3(4):329-354.
[9] Resnick P, Iacovou N, Suchak M, et al. GroupLens: an Open Architecture for CollaborativeFiltering of Netnews[C]. In Proceedings of the1994ACM Conference on Computer SupportedCooperative Work,1994:175-186.
[10] Hill W, Stead L Rosenstein M, et al. Recommending and Evaluating Choices in a VirtualCommunity of Use[C]. In Proceedings of the SIGCHI Conference on Human Factors inComputing Systems,1995:194-201.
[11] Shardanand U, Maes P. Social information filtering algorithms for automating “word ofmouth”[C]. In Proceedings of the SIGCHI Conference on Human Factors in ComputingSystems,1995:210-217.
[12]许海玲,吴潇,李晓东等.互联网推荐系统比较研究[J].软件学报,2009,20(2):350-362.
[13] Malone T W, Grant K R, Turbak F A. The information lens: an intelligent system forinformation sharing in organizations[C]. In Proceedings of the SIGCHI Conference onHuman Factors in Computing Systems,1986:1-8.
[14] Balabanovic M, Shoham Y. Learning Information Retrieval Agents: Experiments withAutomated Web Browsing[C]. In Proceedings of the AAAI Spring Symposium onInformation Gathering,1995:13-18.
[15] Pazzani M, Muramatsu J, Billsus D. Syskill&Webert: Identifying interesting web sites[C]. InProceedings of the13th National Conference on Artificial Intelligence,1996:54-61.
[16] Joachims T, Freitag D, Mitchell T. WebWatcher: A Tour Guide for the World Wide Web[C]. InProceedings of the15th International Joint Conference on Artificial Intellignece,1997:770-775.
[17] ZhangY, Callan J, Minka T. Novelty and redundancy detection in adaptive filtering[C]. InProceedings of the25th Annual International ACM SIGIR Conference on Research andDevelopment in Information Retrieval,2002:81-88.
[18] Degemmis M, Lops P, Semeraro G. A Content-Collaborative Recommender that ExploitsWordNet-based User Profiles for Neighborhood Formation [J]. User Modeling andUser-Adapted Interaction,2007,17(3):217-255.
[19]田超,覃左言,朱青等.SuperRank：基于评论分析的智能推荐系统[J].计算机研究与发展.2010,47(1):494-498.
[20] Chang Y I, Shen J H, Chen T I. A Data Mining-Based Method for the Incremental Update ofSupporting Personalized Information Filtering[J]. Journal of Information Science andEngineering,2008,24(1):129-142.
[21] Goldbery D, Nichols D, Oki B M, et al. Using Collaborative Filtering to Weave anInformation Tapestry[J]. Communications of the ACM,1992,35(12):65-70.
[22] Konstan J A, Miller B N, Maltz D, et al. GroupLens: Applying Collaborative Filtering toUsenet News [J]. Communications of the ACM,1997,40(3):77-87.
[23] Goldberg K, Roeder T, Gupta D, et al. Eigentaste: A Constant Time Collaborative FilteringAlgorithm [J]. Information Retrieval,2001,4(2):133–151.
[24] Breese J S, Heckerman D, Kadie C. Empirical Analysis of Predictive Algorithms forCollaborative Filtering[C]. In Proceedings of the14th Conference on Uncertainty in ArtificialIintelligence,1998:43-52.
[25] Getoor L, Sahami M. Using Probabilistic Relational Models for Collaborative Filtering[C]. InProceedings of the Workshop on Web Usage Analysis and User Profiling,1999.
[26] Sarwar B, Karypis G, Konstan J, et al. Item-Based Collaborative Filtering RecommendationAlgorithms[C]. In Proceedings of the10th International Conference on World Wide Web,2001:285-295.
[27] Pavlov D Y, Pennock D M. A Maximum Entropy Approach to Collaborative Filtering inDynamic, Sparse, High-Dimensional Domains[J]. Advances in Neural Information ProcessingSystems,2002:1441-1448.
[28] Shani G, Brafman R I, Heckerman D. An MDP-based Recommender System[C]. InProceedings of the18th Conference on Uncertainty in Artificial Intelligence,2002:453-460.
[29] Xue G R, Lin C, Yang Q, et al. Scalable Collaborative Filtering using Cluster-BasedSmoothing[C]. In Proceedings of the28th Annual International ACM SIGIR Conference onResearch and Development in Information Retrieval. ACM,2005:114-121.
[30] Das A S, Datar M, Garg A, et al. Google news personalization: Scalable Online CollaborativeFiltering[C]. In Proceedings of the16th International Conference on World Wide Web. ACM,2007:271-280.
[31] Hannon J, Bennett M, Smyth B. Recommending Twitter Users to Follow Using Content andCollaborative Filtering Approaches [C]. In Proceedings of the4th ACM Conference onRecommender Systems. ACM,2010:199-206.
[32] Tsai C F, Hung C. Cluster Ensembles in Collaborative Filtering Recommendation [J]. AppliedSoft Computing,2012,12(4):1417-1425.
[33]赵琴琴,鲁凯,王斌. SPCF:一种基于内存的传播式协同过滤推荐算法[J].计算机学报,2013,36(3):671-676.
[34]贾冬艳,张付志.基于双重邻居选取策略的协同过滤推荐算法[J].计算机研究与发展,2013,50(5):1076-1084.
[35]杨兴耀,于炯,吐尔根等.融合奇异性和扩散过程的协同过滤模型[J].软件学报,2013,24(8):1868-1884.
[36] Marlin B. Collaborative filtering: A machine learning perspective [D]. University of Toronto,2004.
[37] Sarwar B, Karypis G, Konstan J, et al. Item-based collaborative filtering recommendationalgorithms[C]. In Proceedings of the10th international conference on World Wide Web. ACM,2001:285-295.
[38] Yu K, Schwaighofer A, Tresp V, et al. Probabilistic memory-based collaborative filtering [J].IEEE Transactions on Knowledge and Data Engineering,2004,16(1):56-69.
[39] Ungar L H, Foster D P. Clustering methods for collaborative filtering[C]. In AAAI Workshopon Recommendation Systems.1998,1.
[40] Kohrs A, Merialdo B. Clustering for collaborative filtering applications[C].In ComputationalIntelligence for Modelling, Control&Automation. IOS.1999.
[41] Paterek A. Improving regularized singular value decomposition for collaborative filtering[C].In Proceedings of KDD cup and workshop.2007:5-8.
[42] Chien Y H, George E I. A bayesian model for collaborative filtering[C]. In Proceedings of the7th International Workshop on Artificial Intelligence and Statistics. San Francisco: MorganKaufman Publishers,1999.
[43] Marlin B M. Modeling User Rating Profiles For Collaborative Filtering[J]. Advances inNeural Information Processing Systems.2003.
[44] Mirza B J, Keller B J, Ramakrishnan N. Studying recommendation algorithms by graphanalysis [J]. Journal of Intelligent Information Systems,2003,20(2):131-160.
[45] Savia E, Puolam ki K, Kaski S. Latent grouping models for user preference prediction [J].Machine Learning,2009,74(1):75-109.
[46] Salakhutdinov R, Mnih A. Probabilistic Matrix Factorization [J]. Advances in NeuralInformation Processing Systems,2007,1(1):1-8.
[47] Bell R M, Koren Y. Lessons from the Netflix prize challenge [J]. ACM SIGKDD ExplorationsNewsletter,2007,9(2):75-79.
[48] Salakhutdinov R, Mnih A. Bayesian probabilistic matrix factorization using Markov chainMonte Carlo[C]. In Proceedings of the25th international conference on Machine learning.ACM,2008:880-887.
[49] Koren Y, Bell R, Volinsky C. Matrix factorization techniques for recommender systems [J].Computer,2009,42(8):30-37.
[50] Girardi R, Marinho L B. A domain model of Web recommender systems based on usagemining and collaborative filtering [J]. Requirements Engineering,2007,12(1):23-40.
[51] Yoshii K, Goto M, Komatani K, et al. An efficient hybrid music recommender system usingan incrementally trainable probabilistic generative model [J]. IEEE Transactions on Audio,Speech, and Language Processing,2008,16(2):435-447.
[52] Velasquez J D, Palade V. Building a knowledge base for implementing a web-basedcomputerized recommendation system [J]. International Journal on Artificial IntelligenceTools,2007,16(05):793-828.
[53] Aciar S, Zhang D, Simoff S, et al. Informed recommender: Basing recommendations onconsumer product reviews [J]. IEEE Intelligent Systems,2007,22(3):39-47.
[54] Wang H C, Chang Y L. PKR: A personalized knowledge recommendation system for virtualresearch communities [J]. Journal of Computer Information Systems,2007,48(1):31-41.
[55] Agrawal R, Imieliński T, Swami A. Mining association rules between sets of items in largedatabases[C].InACM SIGMOD Record.1993,22(2):207-216.
[56] Han J, Pei J, Yin Y. Mining frequent patterns without candidate generation[C]. In ACMSIGMOD Record.2000,29(2):1-12.
[57] Wang J C, Chiu C C. Recommending trusted online auction sellers using social networkanalysis [J]. Expert Systems with Applications,2008,34(3):1666-1679.
[58] Moon S, Russell G J. Predicting product purchase from inferred customer similarity: Anautologistic model approach [J]. Management Science,2008,54(1):71-82.
[59]王立才,孟祥武,张玉洁.上下文感知推荐系统[J].软件学报,2012,23(1):1-20.
[60]孟祥武,胡勋,王立才等.移动推荐系统及其应用[J].软件学报,2013,24(1):91-108.
[61]郭磊,马军,陈竹敏.一种结合推荐对象间关联关系的社会化推荐算法[J].计算机学报,2014,37(1):219-228.
[62] Mitchell T M著.曾华军,张银奎等译.机器学习[M].北京:机械工业出版社,2011.
[63] Scudder III H. Probability of Error of Some Adaptive Pattern-Recognition Machines [J]. IEEETransactions on Information Theory,1965,11(3):363–371.
[64] Fralick S. Learning to Recognize Patterns without a Teacher [J]. IEEE Transactions onInformation Theory,1967,13(1):57-64.
[65] Agrawala A. Learning with a probabilistic teacher [J]. IEEE Transactions on InformationTheory,1970,16(4):373-379.
[66] Yarowsky D. Unsupervised Word Sense Disambiguation Rivaling Supervised Methods[C]. InProceedings of the33rd Annual Meeting on Association for Computational Linguistics,1995:189-196.
[67] Riloff E, Wiebe J, Wilson T. Learning Subjective Nouns using Extraction PatternBootstrapping [C]. In Proceedings of7th Conference on Natural Language Learning(CoNLL2003),2003:25-32.
[68] Rosenberg C, Hebert M, Schneiderman H. Semi-Supervised Self-Training of ObjectDetection Models [C]. In Proceedings of7th IEEE Workshop on Applications of ComputerVision,2005.
[69] Haffari G R, Sarkar A. Analysis of Semi-Supervised Learning with the Yarowsky Algorithm
[C]. In Proceedings of23rd Conference on Uncertainty in Artificial Intelligence,2007:1-15.
[70] Clup M, Michailidis G. An Iterative Algorithm for Extending Learners to a Semi-SupervisedSetting [J]. Journal of Computational and Graphical Statistics,2008,17(3):545-571.
[71] Baluja S. Probabilistic Modeling for Face Orientation Discrimination: Learning from Labeledand Unlabeled Data [C]. In Proceedings of12th Annual Conference on Neural InformationProcessing Systems (NIPS),1998.
[72] Nigam K, McCallum A K, Thrun S, et al. Text Classification from Labeled and UnlabeledDocuments Using EM [J]. Machine Learning,2000,39(2-3):103-134.
[73] Fujino A, Ueda N, Saito K. A Hybrid Generative/Discriminative Approach to Semi-supervisedClassifier Design [C]. In Proceedings of20th National Conference on Artificial Intelligence,2005:764-769.
[74] Blum A, Mitchell T. Combining Labeled and Unlabeled Data with Co-training [C]. InProceedings of the11th Annual Conference on Computational Learning Theory,1998:92-100.
[75] Goldman S, Zhou Y. Enhancing Supervised Learning with Unlabeled Data [C]. InProceedings of the17th International Conference on Machine Learning,2000:327-334.
[76] Zhou Y, Goldman S. Democratic Co-learning [C]. In Proceedings of the16th IEEEInternational Conference on Tools with Artificial Intelligence,2004:594-602.
[77] Zhou Z H, Li M. Tri-training: Exploiting Unlabeled Data Using Three Classifiers [J]. IEEETransactions on Knowledge and Data Engineering,2005,17(11):1529–1541.
[78] Balcan M F, Blum A, Yang K. Co-training and Expansion: Towards Bridging Theory andPractice [J]. Advances in neural information processing systems,2004,89-96.
[79] Ando R K, Zhang T. Two-View Feature Generation Model for Semi-Supervised Learning [C].In Proceedings of the24th International Conference on Machine Learning,2007,25-32.
[80] Bennett k, Demiriz A. Semi-Supervised Support Vector Machines [J]. Advances in NeuralInformation Processing Systems,1999:368-374.
[81] Joachims T. Transductive Inference for Text ClassificationUsingSupportVector Machines [C].In Proceedings of the16th International Conf. on Machine Learning,1999:200-209.
[82] Bie T D, Cristianini N. Convex Methods for Transduction [J]. Advances in NeuralInformation Processing Systems,2004,16,73–80.
[83] Xu, L, Schuurmans D. Unsupervised and Semi-Supervised Multi-Class Support VectorMachines [C]. In Proceedings of the20th National Conference on Artificial Intelligence,2005.
[84] Chapelle O, Zien A. Semi-Supervised Classification by Low Density Separation [C]. InProceedings of the10th International Workshop on Artificial Intelligence and Statistics,2005.
[85] Sindhwani V, Keerthi S, Chapelle O. Deterministic Annealing for Semi-Supervised KernelMachines [C]. In Proceedings of the23th International Conference on Machine Learning,2006.
[86] Chapelle O, Chi M, Zien A. A Continuation Method for Semi-Supervised SVMs [C]. InProceedings of the23th International Conference on Machine Learning,2006:185-192.
[87] Sindhwani V, Keerthi S S. Large Scale Semi-Supervised Linear SVMs [C]. In Proceedings ofthe29th Annual International ACM SIGIR Conference on Research and Development inInformation Retrieval,2006.
[88] Chapelle O, Sindhwani V, Keerthi S. Branch and Bound for Semi-Supervised Support VectorMachines [J]. Advances in Neural Information Processing Systems,2007.
[89] Lawrence N D, Jordan M I. Semi-Supervised Learning Via Gaussian Processes [C].Advances in Neural Information Processing Systems,2004,17:753-760.
[90] Grandvalet Y, Bengio Y. Semi-Supervised Learning by Entropy Minimization [J].Advances inNeural Information Processing Systems,2005,17:529-536.
[91] Blum A, Chawla S. Learning from Labeled and Unlabeled Data Using Graph Mincuts [C]. InProceedings of the18th International Conference on Machine Learning,2001:19-26.
[92] Zhu X, Ghahramani Z, Lafferty J. Semi-Supervised Learning Using Gaussian Fields andHarmonic Functions [C]. In Proceedings of the20th International Conference on MachineLearning,2003,3:912-919.
[93] Zhou D, Bousquet O, Lal T N, et al. Learning with Local and Global Consistency [J]. InAdvances in Neural Information Processing System,2004,16:321-328.
[94] Belkin M, Matveeva I, Niyogi P. Regularization and Semi-Supervised Learning on LargeGraphs [J]. Lecture Notes in Computer Science,2004,3120:624-638.
[95] Belkin M, Niyogi P, Sindhwani V. Manifold Regularization: A Geometric Framework forLearning from Labeled and Unlabeled Examples [J]. Journal of Machine Learning Research,2006,7:2399-2434.
[96] Vig J, Sen S, Riedl J. Tagsplanations: explaining recommendations using tags [C]. InProceedings of the14th International Conference on Intelligent user interfaces. ACM,2009:47-56.
[97] Shani G, Gunawardana A. Evaluating recommendation systems [M]. Recommender systemshandbook. Springer US,2011:257-297.
[98] Jain A K, Murty M N, Flynn P J. Data Clustering: a Review [J]. ACM Computing Surveys(CSUR),1999,31(3):264-323.
[99] Xu R, Wunsch D. Survey of Clustering Algorithms [J]. IEEE Transactions on NeuralNetworks,2005,16(3):645-678.
[100] MacQueen J. Some Methods for Classification and Analysis of Multivariate Observations [C].In Proceedings of the5th Berkeley Symposium on Mathematical Statistics and Probability,1967,1967,1(281-297):14.
[101] Dempster A P, Laird N M, Rubin D B. Maximum Likelihood from Incomplete Data via theEM algorithm [J]. Journal of the Royal Statistical Society. Series B,1977,39(1):1–38.
[102] Figueiredo M A T, Jain A K, Unsupervised Learning of Finite Mixture Models [J]. IEEETransactions on Pattern Analysis and Machine Intelligence,2002,24(3):381–396.
[103] Ng A Y, Jordan M I, Weiss Y. On Spectral Clustering: Analysis and an Algorithm [J].Advances in Neural Information Processing Systems,2002,2:849–856.
[104] Von Luxburg U. A Tutorial on Spectral Clustering [J]. Statistics and Computing,2007,17(4):395–416.
[105] Xu L, Neufeld J, Larson B, et al. Maximum Margin Clustering [J]. Advances in NeuralInformation Processing Systems,2004,17:1537–1544.
[106] Zhong S. Semi-Supervised Model-Based Document Clustering: A Comparative Study [J].Machine Learning,2006,65(1):3-29.
[107] Bilenko M, Basu S, Mooney R J. Integrating Constraints and Metric Learning inSemi-Supervised Clustering[C]. In Proceedings of the21th International Conference onMachine Learning. ACM,2004:81-88.
[108] Basu S. Semi-Supervised Clustering: Probabilistic Models, Algorithms and Experiments [M].University of Texas at Austin,2005.
[109] Demiriz A, Bennett K P, Embrechts M J. Semi-Supervised Clustering using GeneticAlgorithms [J]. Artificial Neural Networks in Engineering,1999:809-814.
[110] Wagstaff K, Cardie C, Rogers S, et al. Constrained K-Means Clustering with BackgroundKnowledge [C]. In Proceedings of the18th International Conference on Machine Learning,2001,1:577-584.
[111] Basu S, Banerjee A, Mooney R J. Semi-Supervised Clustering by Seeding [C]. In Proceedingsof the19th International Conference on Machine Learning,2002,2:27-34.
[112] Basu S, Bilenko M, Mooney R J. A Probabilistic Framework for Semi-Supervised Clustering
[C]. In Proceedings of the10th ACM SIGKDD International Conference on KnowledgeDiscovery and Data Mining. ACM,2004:59-68.
[113] Bar-Hillel A, Hertz T, Shental N, et al. Learning Distance Functions using EquivalenceRelations [C]. In Proceedings of the20th International Conference on Machine Learning,2003,3:11-18.
[114] Yeung D Y, Chang H. Extending the Relevant Component Analysis Algorithm for MetricLearning using both Positive and Negative Equivalence Constraints[J]. Pattern Recognition,2006,39(5):1007-1010.
[115] Yin X S, Chen S, Hu E, et al. Semi-Supervised Clustering with Metric Learning: an AdaptiveKernel Method [J]. Pattern Recognition,2010,43(4):1320–1333.
[116] Grira N, Crucianu M, Boujemaa N. Unsupervised and Semi-supervised Clustering: a BriefSurvey, In A Review of Machine Learning Techniques for Processing Multimedia Content,Report of the MUSCLE European Network of Excellence (6th Framework Programme),2004.
[117] Ruiz C, Spiliopoulou M, Menasalvas E. Density-based Semi-Supervised Clustering[J]. Datamining and knowledge discovery,2010,21(3):345-370.
[118] Chang C C, Chen H Y. Semi-supervised Clustering with Discriminative Random Fields [J].Pattern Recognition,2012,45(12):4402-4413.
[119] Wagstaff K, Cardie C. Clustering with Instance-level Constraints [C]. In Proceedings of the17th International Conference on Machine Learning,2000:1103–1110.
[120] Cohn D, Caruana R, McCallum A. Semi-Supervised Clustering with User Feedback [J].Constrained Clustering: Advances in Algorithms, Theory, and Applications,2003,4(1):17-32.
[121] Klein D, Kamvar S D, Manning C D. From Instance-level Constraints to Space-LevelConstraints: Making the Most of Prior Knowledge in Data Clustering [C]. In Proceedings ofthe19th International Conference on Machine Learning,2002:307–314.
[122] Xing E P, Ng A Y, Jordan M I, et al. Distance Metric Learning with Application toClustering with Side-information[C]. In Proceedings of the Conference on Advances inNeural Information Processing Systems,2002:505–512.
[123] Cai D, He X F, Han J W. Locally Consistent Concept Factorization for Document Clustering[J]. IEEE Transactions on Knowledge and Data Engineering,2011,23(6):902-913.
[124] He X F, Cai D, Shao Y L, et al. Laplacian regularized Gaussian Mixture Model for DataClustering [J]. IEEE Transactions on Knowledge and Data Engineering,2011,23(9):1406-1418.
[125] Witten I H, Frank E. Data mining: Practical machine learning tools and technique[EB/OL].http://prdownloads.sourceforge.net/weka/datasets-UCI.jar
[126] MovieLens Datasets [EB/OL]. http://grouplens.org/datasets/movielens/
[127] Zhao Y, Kapypis G. Hierarchical Clustering Algorithms for Document Datasets [J], DataMining and Knowledge Discovery,2005,10(2):141–168.
[128] Theobald M,The Software of SVM-light[EB/OL].http://www.mpi-inf.mpg.de/~mtb/svmlight/JNI_SVM-light-6.01.zip
[129] Chapelle O, Scholkopf B, Zien A. Semi-Supervised Learning [M]. Cambridge, MA: MITPress,2006.
[130] Zhu X J. Semi-Supervised Learning Literature Survey [R]. Computer Sciences TR1530,University of Wisconsin at Madison, July19,2008.
[131] Zhang M L, Zhou Z H. COTRADE: Confident Co-Training with Data Editing [J]. IEEETransaction on Systems, Man, and Cybernetics-Part B: Cybernetics,2011,41(6):1612-1626.
[132] Mihalcea R. Co-training and self-training for word sense disambiguation [C]. InProceedings of the Conference on Computational Natural Language Learning(CoNLL-2004).2004.
[133] Tang F, Brennan S, Zhao Q, et al. Co-Tracking Using Semi-Supervised Support VectorMachines [C]. In Proceedings of IEEE the11th International Conference on ComputerVision,2007:1-8.
[134] Wang W, Zhou Z H. A new analysis of co-training [C]. In Proceedings of the27thInternational Conference on Machine Learning (ICML-10).2010:1135-1142.
[135] Yu S, Krishnapuram B, Rosales R, et al. Bayesian Co-Training [J]. Journal of MachineLearning Research,2011,12:2649-2680.
[136] Sun A, Liu Y, Lim E P. Web Classification of Conceptual Entities using Co-Training [J].Expert Systems with Applications,2011,38(12):14367-14375.
[137] Du J, Ling X, Zhou Z H. When does Co-Training Work in Real Data [J]. IEEE Transactionson Knowledge and Data Engineering,2011,23(5):788-799.
[138] Zhu X J, Lafferty J, Ghahramani Z. Combining Active Learning and Semi-SupervisedLearning Using Gaussian Fields and Harmonic Functions [C]. In Proceedings of theICML-2003Workshop on the Continuum from Labeled to Unlabeled Data, Washington DC,2003.
[139] Yang L, Jin R, Sukthankar R. Bayesian Active Distance Metric Learning [C]. In Proceedingsof the23th Conference on Uncertainty in Artificial Intelligence,2009:442-449.
[140] Lughofer E. Hybrid Active Learning for Reducing the Annotation Effort of Operators inClassification Systems [J]. Pattern Recognition,2012,45(2):884-896.
[141] Li H, Shi Y, Liu Y. Cross-domain Video Concept Detection: A Joint Discriminative andGenerative Active Learning Approach [J]. Expert Systems with Applications,2012,39:12220-12228.
[142]2012KDD Cup Track1Datasets [EB/OL].http://www.kddcup2012.org/c/kddcup2012-track1/data
[143] Feng G, Huang G B, Lin Q P, et al. Error Minimized Extreme Learning Machine WithGrowth of Hidden Nodes and Incremental Learning[J]. IEEE Transactions on NeuralNetworks,2009,20(8):1352-1357.
[144] Zhang R, Bawab Z A, Chan A, et al. Investigations on Ensemble Based Semi-SupervisedAcoustic Model Training [C]. In Proceedings of the9th European Conference on SpeechCommunication and Technology,2005:1677-1680.
[145] Karakoulas G, Salakhutdinov R. Semi-supervised mixture-of-experts classification [C]. InProceedings of the4th IEEE International Conference on Data Mining,2004:138-145.
[146] Miller D J, Uyar H S. A mixture of experts classifier with learning based on both labeledand unlabeled data [J]. Advances in neural information processing systems,1997,9:571-577.
[147] Zhang T, Oles F. The value of unlabeled data for classification problems [C]. In Proceedingsof the Seventeenth International Conference on Machine Learning,2000:1191-1198.
[148] Cozman F G, Cohen I. Unlabeled data can degrade classification performance of generativeclassifiers [C]. In Proceedings of the15th international conference of the Florida ArtificialIntelligence Research Society.2002:327-331.
[149] Cohen I, Cozman F G, Sebe N, et al. Semi-supervised learning of classifiers: theory,algorithm, and their application to human-computer interaction [J]. IEEE Transaction onPattern Analysis Machine Intelligence,2004,26(12):1553-1567.
[150] Zhang Y, Yeung D Y. Semi-supervised Generalized Discriminant Analysis [J]. IEEETransactions on Neural Networks,2011,22(8):1207-1217.
[151] Tanha J, Someren M V, Afsarmanesh H. Disagreement-Based Co-Training[C]. InProceedings of the23rd IEEE International Conference on Tools with Artificial Intelligence,2011:803-810.
[152] Chang C C, Pao H K, Lee Y J. An RSVM based two-teachers–one-student semi-supervisedlearning algorithm [J]. Neural Networks,2012,25:57-69.
[153] Zhang R, Rudnicky A I, A New Data Selection Approach for Semi-Supervised AcousticModeling [C]. In Proceedings of IEEE International Conference on Acoustics, Speech andSignal,2006.
[154] Jeon J H, Liu Y. Automatic prosodic event detection using a novel labeling and selectionmethod in co-training [J]. Speech Communication,2012,54(3):445-458.
[155] Kumar Mallapragada P, Jin R, Jain A K, et al. Semiboost: Boosting for semi-supervisedlearning [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2009,31(11):2000-2014.
[156] Blum A, Lafferty J, Rwebangira M R, et al. Semi-supervised learning using randomizedmincuts [C]. In Proceedings of the21th International Conference on Machine Learning,2004:13-20.
[157] Kveton B, Valko M, Rahimi A, et al. Semi-Supervised Learning with Max-Margin Graphcuts [C]. In Proceedings of the13th International Conference on Artificial Intelligence andStatistics,2010:421-428.
[158] Joachims T. Transductive Learning via Spectral Graph Partitioning [C]. In Proceedings ofthe20th International Conference on Machine Learning,2003.
[159] Zhu X J, Ghahramani Z, Lafferty J. Semi-Supervised Learning Using Gaussian Fields andHarmonic Functions [C]. In Proceedings of the12th International Conference on MachineLearning,2003.
[160] Belkin M, Niyogi P, Sindhwani V. On manifold regularization [C]. In Proceedings ofInternational Conference on Artificial Intelligence and Statistics,2005.
[161] Belkin M, Niyogi P, Sindhwani V. Manifold regularization: a Geometric framework forlearning from labeled and unlabeled examples [J]. Journal of Machine Learning Research,2006,7:2399-2434.
[162] Sindhwani V, Hu J, Mojsilovic A. Regularized Co-Clustering with Dual Supervision [J].Advances in Neural Information Processing Systems,2008:976-983.
[163] Jebara T, Wang J, Chang S F. Graph Construction and b-Matching for Semi-SupervisedLearning[C]. In Proceedings of the26th Annual International Conference on MachineLearning,2009:441-448.
[164] Cheng B, Yang J C, Yan S C, et al. Learning with1–Graph for Image Analysis [J]. IEEETransaction on Image Processing,2010,19(4):858-866.
[165] Kapoor A, Qi Y, Ahn H, et al. Hyperparameter and Kernel Learning for Graph BasedSemi-Supervised Classification [C]. Advances in Neural Information Processing Systems,2005.
[166] Zhang X H, Lee W S. Hyperparameter Learning for Graph Based Semi-SupervisedLearning Algorithms [J]. Advances in Neural Information Processing Systems,2006.
[167] Rohban M H, Rabiee H R. Supervised Neighborhood graph construction forsemi-supervised classification [J]. Pattern Recognition,2012,45:1363-1372.
[168] Sugiyama M, Ide T, Nakajima S, et al. Semi-Supervised Local Fisher Discriminant Analysisfor Dimensionality Reduction [J]. Mach Learn,2010,78:35-61.
[169] Zhang F, Zhang J S. Label Propagation through Sparse Neighborhood and its Applications[J]. Neurocomputing,2012,97:267-277.
[2] Borkar V, Carey M J, Li C. Inside “Big Data management”: Ogres, Onions, or Parfaits?[C]. InProceedings of the15thInternational Conference on Extending Database Technology,2012:3-14.
[3]项亮编,陈义,王益审校.推荐系统实践[M].北京:人民邮电出版社,2013.
[4] Adomavicius G, Tuzhilin A. Towards the Next Generation of Recommender Systems: A Surveyof the State-of-the-Art and Possible Extensions [J]. IEEE Transactions on Knowledge and DataEngineering,2005,17(6):734–749.
[5]吴金龙. Netflix Prize中的协同过滤算法[D].北京:北京大学,2010.
[6] Sun J T, Zeng H J, Liu H, et al. CubeSVD: a Novel Approach to Personalized Web Search[C].In Proceedings of the14th International Conference on World Wide Web,2005:382-390.
[7] Sun J T, Wang X H, Shen D, et al. Mining Click through Data for Collaborative Web Search [C].In Proceedings ofthe15th International Conference on World Wide Web,2006:947-948.
[8] Rich E. User Modeling via Stereotypes [J]. Cognitive Science,1979,3(4):329-354.
[9] Resnick P, Iacovou N, Suchak M, et al. GroupLens: an Open Architecture for CollaborativeFiltering of Netnews[C]. In Proceedings of the1994ACM Conference on Computer SupportedCooperative Work,1994:175-186.
[10] Hill W, Stead L Rosenstein M, et al. Recommending and Evaluating Choices in a VirtualCommunity of Use[C]. In Proceedings of the SIGCHI Conference on Human Factors inComputing Systems,1995:194-201.
[11] Shardanand U, Maes P. Social information filtering algorithms for automating “word ofmouth”[C]. In Proceedings of the SIGCHI Conference on Human Factors in ComputingSystems,1995:210-217.
[12]许海玲,吴潇,李晓东等.互联网推荐系统比较研究[J].软件学报,2009,20(2):350-362.
[13] Malone T W, Grant K R, Turbak F A. The information lens: an intelligent system forinformation sharing in organizations[C]. In Proceedings of the SIGCHI Conference onHuman Factors in Computing Systems,1986:1-8.
[14] Balabanovic M, Shoham Y. Learning Information Retrieval Agents: Experiments withAutomated Web Browsing[C]. In Proceedings of the AAAI Spring Symposium onInformation Gathering,1995:13-18.
[15] Pazzani M, Muramatsu J, Billsus D. Syskill&Webert: Identifying interesting web sites[C]. InProceedings of the13th National Conference on Artificial Intelligence,1996:54-61.
[16] Joachims T, Freitag D, Mitchell T. WebWatcher: A Tour Guide for the World Wide Web[C]. InProceedings of the15th International Joint Conference on Artificial Intellignece,1997:770-775.
[17] ZhangY, Callan J, Minka T. Novelty and redundancy detection in adaptive filtering[C]. InProceedings of the25th Annual International ACM SIGIR Conference on Research andDevelopment in Information Retrieval,2002:81-88.
[18] Degemmis M, Lops P, Semeraro G. A Content-Collaborative Recommender that ExploitsWordNet-based User Profiles for Neighborhood Formation [J]. User Modeling andUser-Adapted Interaction,2007,17(3):217-255.
[19]田超,覃左言,朱青等.SuperRank：基于评论分析的智能推荐系统[J].计算机研究与发展.2010,47(1):494-498.
[20] Chang Y I, Shen J H, Chen T I. A Data Mining-Based Method for the Incremental Update ofSupporting Personalized Information Filtering[J]. Journal of Information Science andEngineering,2008,24(1):129-142.
[21] Goldbery D, Nichols D, Oki B M, et al. Using Collaborative Filtering to Weave anInformation Tapestry[J]. Communications of the ACM,1992,35(12):65-70.
[22] Konstan J A, Miller B N, Maltz D, et al. GroupLens: Applying Collaborative Filtering toUsenet News [J]. Communications of the ACM,1997,40(3):77-87.
[23] Goldberg K, Roeder T, Gupta D, et al. Eigentaste: A Constant Time Collaborative FilteringAlgorithm [J]. Information Retrieval,2001,4(2):133–151.
[24] Breese J S, Heckerman D, Kadie C. Empirical Analysis of Predictive Algorithms forCollaborative Filtering[C]. In Proceedings of the14th Conference on Uncertainty in ArtificialIintelligence,1998:43-52.
[25] Getoor L, Sahami M. Using Probabilistic Relational Models for Collaborative Filtering[C]. InProceedings of the Workshop on Web Usage Analysis and User Profiling,1999.
[26] Sarwar B, Karypis G, Konstan J, et al. Item-Based Collaborative Filtering RecommendationAlgorithms[C]. In Proceedings of the10th International Conference on World Wide Web,2001:285-295.
[27] Pavlov D Y, Pennock D M. A Maximum Entropy Approach to Collaborative Filtering inDynamic, Sparse, High-Dimensional Domains[J]. Advances in Neural Information ProcessingSystems,2002:1441-1448.
[28] Shani G, Brafman R I, Heckerman D. An MDP-based Recommender System[C]. InProceedings of the18th Conference on Uncertainty in Artificial Intelligence,2002:453-460.
[29] Xue G R, Lin C, Yang Q, et al. Scalable Collaborative Filtering using Cluster-BasedSmoothing[C]. In Proceedings of the28th Annual International ACM SIGIR Conference onResearch and Development in Information Retrieval. ACM,2005:114-121.
[30] Das A S, Datar M, Garg A, et al. Google news personalization: Scalable Online CollaborativeFiltering[C]. In Proceedings of the16th International Conference on World Wide Web. ACM,2007:271-280.
[31] Hannon J, Bennett M, Smyth B. Recommending Twitter Users to Follow Using Content andCollaborative Filtering Approaches [C]. In Proceedings of the4th ACM Conference onRecommender Systems. ACM,2010:199-206.
[32] Tsai C F, Hung C. Cluster Ensembles in Collaborative Filtering Recommendation [J]. AppliedSoft Computing,2012,12(4):1417-1425.
[33]赵琴琴,鲁凯,王斌. SPCF:一种基于内存的传播式协同过滤推荐算法[J].计算机学报,2013,36(3):671-676.
[34]贾冬艳,张付志.基于双重邻居选取策略的协同过滤推荐算法[J].计算机研究与发展,2013,50(5):1076-1084.
[35]杨兴耀,于炯,吐尔根等.融合奇异性和扩散过程的协同过滤模型[J].软件学报,2013,24(8):1868-1884.
[36] Marlin B. Collaborative filtering: A machine learning perspective [D]. University of Toronto,2004.
[37] Sarwar B, Karypis G, Konstan J, et al. Item-based collaborative filtering recommendationalgorithms[C]. In Proceedings of the10th international conference on World Wide Web. ACM,2001:285-295.
[38] Yu K, Schwaighofer A, Tresp V, et al. Probabilistic memory-based collaborative filtering [J].IEEE Transactions on Knowledge and Data Engineering,2004,16(1):56-69.
[39] Ungar L H, Foster D P. Clustering methods for collaborative filtering[C]. In AAAI Workshopon Recommendation Systems.1998,1.
[40] Kohrs A, Merialdo B. Clustering for collaborative filtering applications[C].In ComputationalIntelligence for Modelling, Control&Automation. IOS.1999.
[41] Paterek A. Improving regularized singular value decomposition for collaborative filtering[C].In Proceedings of KDD cup and workshop.2007:5-8.
[42] Chien Y H, George E I. A bayesian model for collaborative filtering[C]. In Proceedings of the7th International Workshop on Artificial Intelligence and Statistics. San Francisco: MorganKaufman Publishers,1999.
[43] Marlin B M. Modeling User Rating Profiles For Collaborative Filtering[J]. Advances inNeural Information Processing Systems.2003.
[44] Mirza B J, Keller B J, Ramakrishnan N. Studying recommendation algorithms by graphanalysis [J]. Journal of Intelligent Information Systems,2003,20(2):131-160.
[45] Savia E, Puolam ki K, Kaski S. Latent grouping models for user preference prediction [J].Machine Learning,2009,74(1):75-109.
[46] Salakhutdinov R, Mnih A. Probabilistic Matrix Factorization [J]. Advances in NeuralInformation Processing Systems,2007,1(1):1-8.
[47] Bell R M, Koren Y. Lessons from the Netflix prize challenge [J]. ACM SIGKDD ExplorationsNewsletter,2007,9(2):75-79.
[48] Salakhutdinov R, Mnih A. Bayesian probabilistic matrix factorization using Markov chainMonte Carlo[C]. In Proceedings of the25th international conference on Machine learning.ACM,2008:880-887.
[49] Koren Y, Bell R, Volinsky C. Matrix factorization techniques for recommender systems [J].Computer,2009,42(8):30-37.
[50] Girardi R, Marinho L B. A domain model of Web recommender systems based on usagemining and collaborative filtering [J]. Requirements Engineering,2007,12(1):23-40.
[51] Yoshii K, Goto M, Komatani K, et al. An efficient hybrid music recommender system usingan incrementally trainable probabilistic generative model [J]. IEEE Transactions on Audio,Speech, and Language Processing,2008,16(2):435-447.
[52] Velasquez J D, Palade V. Building a knowledge base for implementing a web-basedcomputerized recommendation system [J]. International Journal on Artificial IntelligenceTools,2007,16(05):793-828.
[53] Aciar S, Zhang D, Simoff S, et al. Informed recommender: Basing recommendations onconsumer product reviews [J]. IEEE Intelligent Systems,2007,22(3):39-47.
[54] Wang H C, Chang Y L. PKR: A personalized knowledge recommendation system for virtualresearch communities [J]. Journal of Computer Information Systems,2007,48(1):31-41.
[55] Agrawal R, Imieliński T, Swami A. Mining association rules between sets of items in largedatabases[C].InACM SIGMOD Record.1993,22(2):207-216.
[56] Han J, Pei J, Yin Y. Mining frequent patterns without candidate generation[C]. In ACMSIGMOD Record.2000,29(2):1-12.
[57] Wang J C, Chiu C C. Recommending trusted online auction sellers using social networkanalysis [J]. Expert Systems with Applications,2008,34(3):1666-1679.
[58] Moon S, Russell G J. Predicting product purchase from inferred customer similarity: Anautologistic model approach [J]. Management Science,2008,54(1):71-82.
[59]王立才,孟祥武,张玉洁.上下文感知推荐系统[J].软件学报,2012,23(1):1-20.
[60]孟祥武,胡勋,王立才等.移动推荐系统及其应用[J].软件学报,2013,24(1):91-108.
[61]郭磊,马军,陈竹敏.一种结合推荐对象间关联关系的社会化推荐算法[J].计算机学报,2014,37(1):219-228.
[62] Mitchell T M著.曾华军,张银奎等译.机器学习[M].北京:机械工业出版社,2011.
[63] Scudder III H. Probability of Error of Some Adaptive Pattern-Recognition Machines [J]. IEEETransactions on Information Theory,1965,11(3):363–371.
[64] Fralick S. Learning to Recognize Patterns without a Teacher [J]. IEEE Transactions onInformation Theory,1967,13(1):57-64.
[65] Agrawala A. Learning with a probabilistic teacher [J]. IEEE Transactions on InformationTheory,1970,16(4):373-379.
[66] Yarowsky D. Unsupervised Word Sense Disambiguation Rivaling Supervised Methods[C]. InProceedings of the33rd Annual Meeting on Association for Computational Linguistics,1995:189-196.
[67] Riloff E, Wiebe J, Wilson T. Learning Subjective Nouns using Extraction PatternBootstrapping [C]. In Proceedings of7th Conference on Natural Language Learning(CoNLL2003),2003:25-32.
[68] Rosenberg C, Hebert M, Schneiderman H. Semi-Supervised Self-Training of ObjectDetection Models [C]. In Proceedings of7th IEEE Workshop on Applications of ComputerVision,2005.
[69] Haffari G R, Sarkar A. Analysis of Semi-Supervised Learning with the Yarowsky Algorithm
[C]. In Proceedings of23rd Conference on Uncertainty in Artificial Intelligence,2007:1-15.
[70] Clup M, Michailidis G. An Iterative Algorithm for Extending Learners to a Semi-SupervisedSetting [J]. Journal of Computational and Graphical Statistics,2008,17(3):545-571.
[71] Baluja S. Probabilistic Modeling for Face Orientation Discrimination: Learning from Labeledand Unlabeled Data [C]. In Proceedings of12th Annual Conference on Neural InformationProcessing Systems (NIPS),1998.
[72] Nigam K, McCallum A K, Thrun S, et al. Text Classification from Labeled and UnlabeledDocuments Using EM [J]. Machine Learning,2000,39(2-3):103-134.
[73] Fujino A, Ueda N, Saito K. A Hybrid Generative/Discriminative Approach to Semi-supervisedClassifier Design [C]. In Proceedings of20th National Conference on Artificial Intelligence,2005:764-769.
[74] Blum A, Mitchell T. Combining Labeled and Unlabeled Data with Co-training [C]. InProceedings of the11th Annual Conference on Computational Learning Theory,1998:92-100.
[75] Goldman S, Zhou Y. Enhancing Supervised Learning with Unlabeled Data [C]. InProceedings of the17th International Conference on Machine Learning,2000:327-334.
[76] Zhou Y, Goldman S. Democratic Co-learning [C]. In Proceedings of the16th IEEEInternational Conference on Tools with Artificial Intelligence,2004:594-602.
[77] Zhou Z H, Li M. Tri-training: Exploiting Unlabeled Data Using Three Classifiers [J]. IEEETransactions on Knowledge and Data Engineering,2005,17(11):1529–1541.
[78] Balcan M F, Blum A, Yang K. Co-training and Expansion: Towards Bridging Theory andPractice [J]. Advances in neural information processing systems,2004,89-96.
[79] Ando R K, Zhang T. Two-View Feature Generation Model for Semi-Supervised Learning [C].In Proceedings of the24th International Conference on Machine Learning,2007,25-32.
[80] Bennett k, Demiriz A. Semi-Supervised Support Vector Machines [J]. Advances in NeuralInformation Processing Systems,1999:368-374.
[81] Joachims T. Transductive Inference for Text ClassificationUsingSupportVector Machines [C].In Proceedings of the16th International Conf. on Machine Learning,1999:200-209.
[82] Bie T D, Cristianini N. Convex Methods for Transduction [J]. Advances in NeuralInformation Processing Systems,2004,16,73–80.
[83] Xu, L, Schuurmans D. Unsupervised and Semi-Supervised Multi-Class Support VectorMachines [C]. In Proceedings of the20th National Conference on Artificial Intelligence,2005.
[84] Chapelle O, Zien A. Semi-Supervised Classification by Low Density Separation [C]. InProceedings of the10th International Workshop on Artificial Intelligence and Statistics,2005.
[85] Sindhwani V, Keerthi S, Chapelle O. Deterministic Annealing for Semi-Supervised KernelMachines [C]. In Proceedings of the23th International Conference on Machine Learning,2006.
[86] Chapelle O, Chi M, Zien A. A Continuation Method for Semi-Supervised SVMs [C]. InProceedings of the23th International Conference on Machine Learning,2006:185-192.
[87] Sindhwani V, Keerthi S S. Large Scale Semi-Supervised Linear SVMs [C]. In Proceedings ofthe29th Annual International ACM SIGIR Conference on Research and Development inInformation Retrieval,2006.
[88] Chapelle O, Sindhwani V, Keerthi S. Branch and Bound for Semi-Supervised Support VectorMachines [J]. Advances in Neural Information Processing Systems,2007.
[89] Lawrence N D, Jordan M I. Semi-Supervised Learning Via Gaussian Processes [C].Advances in Neural Information Processing Systems,2004,17:753-760.
[90] Grandvalet Y, Bengio Y. Semi-Supervised Learning by Entropy Minimization [J].Advances inNeural Information Processing Systems,2005,17:529-536.
[91] Blum A, Chawla S. Learning from Labeled and Unlabeled Data Using Graph Mincuts [C]. InProceedings of the18th International Conference on Machine Learning,2001:19-26.
[92] Zhu X, Ghahramani Z, Lafferty J. Semi-Supervised Learning Using Gaussian Fields andHarmonic Functions [C]. In Proceedings of the20th International Conference on MachineLearning,2003,3:912-919.
[93] Zhou D, Bousquet O, Lal T N, et al. Learning with Local and Global Consistency [J]. InAdvances in Neural Information Processing System,2004,16:321-328.
[94] Belkin M, Matveeva I, Niyogi P. Regularization and Semi-Supervised Learning on LargeGraphs [J]. Lecture Notes in Computer Science,2004,3120:624-638.
[95] Belkin M, Niyogi P, Sindhwani V. Manifold Regularization: A Geometric Framework forLearning from Labeled and Unlabeled Examples [J]. Journal of Machine Learning Research,2006,7:2399-2434.
[96] Vig J, Sen S, Riedl J. Tagsplanations: explaining recommendations using tags [C]. InProceedings of the14th International Conference on Intelligent user interfaces. ACM,2009:47-56.
[97] Shani G, Gunawardana A. Evaluating recommendation systems [M]. Recommender systemshandbook. Springer US,2011:257-297.
[98] Jain A K, Murty M N, Flynn P J. Data Clustering: a Review [J]. ACM Computing Surveys(CSUR),1999,31(3):264-323.
[99] Xu R, Wunsch D. Survey of Clustering Algorithms [J]. IEEE Transactions on NeuralNetworks,2005,16(3):645-678.
[100] MacQueen J. Some Methods for Classification and Analysis of Multivariate Observations [C].In Proceedings of the5th Berkeley Symposium on Mathematical Statistics and Probability,1967,1967,1(281-297):14.
[101] Dempster A P, Laird N M, Rubin D B. Maximum Likelihood from Incomplete Data via theEM algorithm [J]. Journal of the Royal Statistical Society. Series B,1977,39(1):1–38.
[102] Figueiredo M A T, Jain A K, Unsupervised Learning of Finite Mixture Models [J]. IEEETransactions on Pattern Analysis and Machine Intelligence,2002,24(3):381–396.
[103] Ng A Y, Jordan M I, Weiss Y. On Spectral Clustering: Analysis and an Algorithm [J].Advances in Neural Information Processing Systems,2002,2:849–856.
[104] Von Luxburg U. A Tutorial on Spectral Clustering [J]. Statistics and Computing,2007,17(4):395–416.
[105] Xu L, Neufeld J, Larson B, et al. Maximum Margin Clustering [J]. Advances in NeuralInformation Processing Systems,2004,17:1537–1544.
[106] Zhong S. Semi-Supervised Model-Based Document Clustering: A Comparative Study [J].Machine Learning,2006,65(1):3-29.
[107] Bilenko M, Basu S, Mooney R J. Integrating Constraints and Metric Learning inSemi-Supervised Clustering[C]. In Proceedings of the21th International Conference onMachine Learning. ACM,2004:81-88.
[108] Basu S. Semi-Supervised Clustering: Probabilistic Models, Algorithms and Experiments [M].University of Texas at Austin,2005.
[109] Demiriz A, Bennett K P, Embrechts M J. Semi-Supervised Clustering using GeneticAlgorithms [J]. Artificial Neural Networks in Engineering,1999:809-814.
[110] Wagstaff K, Cardie C, Rogers S, et al. Constrained K-Means Clustering with BackgroundKnowledge [C]. In Proceedings of the18th International Conference on Machine Learning,2001,1:577-584.
[111] Basu S, Banerjee A, Mooney R J. Semi-Supervised Clustering by Seeding [C]. In Proceedingsof the19th International Conference on Machine Learning,2002,2:27-34.
[112] Basu S, Bilenko M, Mooney R J. A Probabilistic Framework for Semi-Supervised Clustering
[C]. In Proceedings of the10th ACM SIGKDD International Conference on KnowledgeDiscovery and Data Mining. ACM,2004:59-68.
[113] Bar-Hillel A, Hertz T, Shental N, et al. Learning Distance Functions using EquivalenceRelations [C]. In Proceedings of the20th International Conference on Machine Learning,2003,3:11-18.
[114] Yeung D Y, Chang H. Extending the Relevant Component Analysis Algorithm for MetricLearning using both Positive and Negative Equivalence Constraints[J]. Pattern Recognition,2006,39(5):1007-1010.
[115] Yin X S, Chen S, Hu E, et al. Semi-Supervised Clustering with Metric Learning: an AdaptiveKernel Method [J]. Pattern Recognition,2010,43(4):1320–1333.
[116] Grira N, Crucianu M, Boujemaa N. Unsupervised and Semi-supervised Clustering: a BriefSurvey, In A Review of Machine Learning Techniques for Processing Multimedia Content,Report of the MUSCLE European Network of Excellence (6th Framework Programme),2004.
[117] Ruiz C, Spiliopoulou M, Menasalvas E. Density-based Semi-Supervised Clustering[J]. Datamining and knowledge discovery,2010,21(3):345-370.
[118] Chang C C, Chen H Y. Semi-supervised Clustering with Discriminative Random Fields [J].Pattern Recognition,2012,45(12):4402-4413.
[119] Wagstaff K, Cardie C. Clustering with Instance-level Constraints [C]. In Proceedings of the17th International Conference on Machine Learning,2000:1103–1110.
[120] Cohn D, Caruana R, McCallum A. Semi-Supervised Clustering with User Feedback [J].Constrained Clustering: Advances in Algorithms, Theory, and Applications,2003,4(1):17-32.
[121] Klein D, Kamvar S D, Manning C D. From Instance-level Constraints to Space-LevelConstraints: Making the Most of Prior Knowledge in Data Clustering [C]. In Proceedings ofthe19th International Conference on Machine Learning,2002:307–314.
[122] Xing E P, Ng A Y, Jordan M I, et al. Distance Metric Learning with Application toClustering with Side-information[C]. In Proceedings of the Conference on Advances inNeural Information Processing Systems,2002:505–512.
[123] Cai D, He X F, Han J W. Locally Consistent Concept Factorization for Document Clustering[J]. IEEE Transactions on Knowledge and Data Engineering,2011,23(6):902-913.
[124] He X F, Cai D, Shao Y L, et al. Laplacian regularized Gaussian Mixture Model for DataClustering [J]. IEEE Transactions on Knowledge and Data Engineering,2011,23(9):1406-1418.
[125] Witten I H, Frank E. Data mining: Practical machine learning tools and technique[EB/OL].http://prdownloads.sourceforge.net/weka/datasets-UCI.jar
[126] MovieLens Datasets [EB/OL]. http://grouplens.org/datasets/movielens/
[127] Zhao Y, Kapypis G. Hierarchical Clustering Algorithms for Document Datasets [J], DataMining and Knowledge Discovery,2005,10(2):141–168.
[128] Theobald M,The Software of SVM-light[EB/OL].http://www.mpi-inf.mpg.de/~mtb/svmlight/JNI_SVM-light-6.01.zip
[129] Chapelle O, Scholkopf B, Zien A. Semi-Supervised Learning [M]. Cambridge, MA: MITPress,2006.
[130] Zhu X J. Semi-Supervised Learning Literature Survey [R]. Computer Sciences TR1530,University of Wisconsin at Madison, July19,2008.
[131] Zhang M L, Zhou Z H. COTRADE: Confident Co-Training with Data Editing [J]. IEEETransaction on Systems, Man, and Cybernetics-Part B: Cybernetics,2011,41(6):1612-1626.
[132] Mihalcea R. Co-training and self-training for word sense disambiguation [C]. InProceedings of the Conference on Computational Natural Language Learning(CoNLL-2004).2004.
[133] Tang F, Brennan S, Zhao Q, et al. Co-Tracking Using Semi-Supervised Support VectorMachines [C]. In Proceedings of IEEE the11th International Conference on ComputerVision,2007:1-8.
[134] Wang W, Zhou Z H. A new analysis of co-training [C]. In Proceedings of the27thInternational Conference on Machine Learning (ICML-10).2010:1135-1142.
[135] Yu S, Krishnapuram B, Rosales R, et al. Bayesian Co-Training [J]. Journal of MachineLearning Research,2011,12:2649-2680.
[136] Sun A, Liu Y, Lim E P. Web Classification of Conceptual Entities using Co-Training [J].Expert Systems with Applications,2011,38(12):14367-14375.
[137] Du J, Ling X, Zhou Z H. When does Co-Training Work in Real Data [J]. IEEE Transactionson Knowledge and Data Engineering,2011,23(5):788-799.
[138] Zhu X J, Lafferty J, Ghahramani Z. Combining Active Learning and Semi-SupervisedLearning Using Gaussian Fields and Harmonic Functions [C]. In Proceedings of theICML-2003Workshop on the Continuum from Labeled to Unlabeled Data, Washington DC,2003.
[139] Yang L, Jin R, Sukthankar R. Bayesian Active Distance Metric Learning [C]. In Proceedingsof the23th Conference on Uncertainty in Artificial Intelligence,2009:442-449.
[140] Lughofer E. Hybrid Active Learning for Reducing the Annotation Effort of Operators inClassification Systems [J]. Pattern Recognition,2012,45(2):884-896.
[141] Li H, Shi Y, Liu Y. Cross-domain Video Concept Detection: A Joint Discriminative andGenerative Active Learning Approach [J]. Expert Systems with Applications,2012,39:12220-12228.
[142]2012KDD Cup Track1Datasets [EB/OL].http://www.kddcup2012.org/c/kddcup2012-track1/data
[143] Feng G, Huang G B, Lin Q P, et al. Error Minimized Extreme Learning Machine WithGrowth of Hidden Nodes and Incremental Learning[J]. IEEE Transactions on NeuralNetworks,2009,20(8):1352-1357.
[144] Zhang R, Bawab Z A, Chan A, et al. Investigations on Ensemble Based Semi-SupervisedAcoustic Model Training [C]. In Proceedings of the9th European Conference on SpeechCommunication and Technology,2005:1677-1680.
[145] Karakoulas G, Salakhutdinov R. Semi-supervised mixture-of-experts classification [C]. InProceedings of the4th IEEE International Conference on Data Mining,2004:138-145.
[146] Miller D J, Uyar H S. A mixture of experts classifier with learning based on both labeledand unlabeled data [J]. Advances in neural information processing systems,1997,9:571-577.
[147] Zhang T, Oles F. The value of unlabeled data for classification problems [C]. In Proceedingsof the Seventeenth International Conference on Machine Learning,2000:1191-1198.
[148] Cozman F G, Cohen I. Unlabeled data can degrade classification performance of generativeclassifiers [C]. In Proceedings of the15th international conference of the Florida ArtificialIntelligence Research Society.2002:327-331.
[149] Cohen I, Cozman F G, Sebe N, et al. Semi-supervised learning of classifiers: theory,algorithm, and their application to human-computer interaction [J]. IEEE Transaction onPattern Analysis Machine Intelligence,2004,26(12):1553-1567.
[150] Zhang Y, Yeung D Y. Semi-supervised Generalized Discriminant Analysis [J]. IEEETransactions on Neural Networks,2011,22(8):1207-1217.
[151] Tanha J, Someren M V, Afsarmanesh H. Disagreement-Based Co-Training[C]. InProceedings of the23rd IEEE International Conference on Tools with Artificial Intelligence,2011:803-810.
[152] Chang C C, Pao H K, Lee Y J. An RSVM based two-teachers–one-student semi-supervisedlearning algorithm [J]. Neural Networks,2012,25:57-69.
[153] Zhang R, Rudnicky A I, A New Data Selection Approach for Semi-Supervised AcousticModeling [C]. In Proceedings of IEEE International Conference on Acoustics, Speech andSignal,2006.
[154] Jeon J H, Liu Y. Automatic prosodic event detection using a novel labeling and selectionmethod in co-training [J]. Speech Communication,2012,54(3):445-458.
[155] Kumar Mallapragada P, Jin R, Jain A K, et al. Semiboost: Boosting for semi-supervisedlearning [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2009,31(11):2000-2014.
[156] Blum A, Lafferty J, Rwebangira M R, et al. Semi-supervised learning using randomizedmincuts [C]. In Proceedings of the21th International Conference on Machine Learning,2004:13-20.
[157] Kveton B, Valko M, Rahimi A, et al. Semi-Supervised Learning with Max-Margin Graphcuts [C]. In Proceedings of the13th International Conference on Artificial Intelligence andStatistics,2010:421-428.
[158] Joachims T. Transductive Learning via Spectral Graph Partitioning [C]. In Proceedings ofthe20th International Conference on Machine Learning,2003.
[159] Zhu X J, Ghahramani Z, Lafferty J. Semi-Supervised Learning Using Gaussian Fields andHarmonic Functions [C]. In Proceedings of the12th International Conference on MachineLearning,2003.
[160] Belkin M, Niyogi P, Sindhwani V. On manifold regularization [C]. In Proceedings ofInternational Conference on Artificial Intelligence and Statistics,2005.
[161] Belkin M, Niyogi P, Sindhwani V. Manifold regularization: a Geometric framework forlearning from labeled and unlabeled examples [J]. Journal of Machine Learning Research,2006,7:2399-2434.
[162] Sindhwani V, Hu J, Mojsilovic A. Regularized Co-Clustering with Dual Supervision [J].Advances in Neural Information Processing Systems,2008:976-983.
[163] Jebara T, Wang J, Chang S F. Graph Construction and b-Matching for Semi-SupervisedLearning[C]. In Proceedings of the26th Annual International Conference on MachineLearning,2009:441-448.
[164] Cheng B, Yang J C, Yan S C, et al. Learning with1–Graph for Image Analysis [J]. IEEETransaction on Image Processing,2010,19(4):858-866.
[165] Kapoor A, Qi Y, Ahn H, et al. Hyperparameter and Kernel Learning for Graph BasedSemi-Supervised Classification [C]. Advances in Neural Information Processing Systems,2005.
[166] Zhang X H, Lee W S. Hyperparameter Learning for Graph Based Semi-SupervisedLearning Algorithms [J]. Advances in Neural Information Processing Systems,2006.
[167] Rohban M H, Rabiee H R. Supervised Neighborhood graph construction forsemi-supervised classification [J]. Pattern Recognition,2012,45:1363-1372.
[168] Sugiyama M, Ide T, Nakajima S, et al. Semi-Supervised Local Fisher Discriminant Analysisfor Dimensionality Reduction [J]. Mach Learn,2010,78:35-61.
[169] Zhang F, Zhang J S. Label Propagation through Sparse Neighborhood and its Applications[J]. Neurocomputing,2012,97:267-277.