In this article, we address the issue of tag recommendation from a machine learning perspective. From our empirical observation of two large-scale datasets, we first argue that the user-centered approach for tag recommendation is not very effective in practice. Consequently, we propose two novel document-centered approaches that are capable of making effective and efficient tag recommendations in real scenarios. The first, graph-based, method represents the tagged data in two bipartite graphs, (document, tag) and (document, word), then finds document topics by leveraging graph partitioning algorithms. The second, prototype-based, method aims at finding the most representative documents within the data collections and advocates a sparse multiclass Gaussian process classifier for efficient document classification. For both methods, tags are ranked within each topic cluster/class by a novel ranking method. Recommendations are performed by first classifying a new document into one or more topic clusters/classes, and then selecting the most relevant tags from those clusters/classes as machine-recommended tags.
Experiments on real-world data from Del.icio.us, CiteULike, and BibSonomy examine the quality of tag recommendation as well as the efficiency of our recommendation algorithms. The results suggest that our document-centered models can substantially improve the performance of tag recommendations when compared to the user-centered methods, as well as topic models LDA and SVM classifiers.}, address = {New York, NY, USA}, author = {Song, Yang and Zhang, Lu and Giles, C. Lee}, doi = {10.1145/1921591.1921595}, interhash = {6e93d08c935eaf887ed46750f405e742}, intrahash = {8e5fdf385f7bae639ca978259d9ec8de}, issn = {1559-1131}, journal = {Transactions on the Web}, month = feb, number = 1, pages = {1--31}, publisher = {ACM}, title = {Automatic tag recommendation algorithms for social recommender systems}, url = {http://doi.acm.org/10.1145/1921591.1921595}, volume = 5, year = 2011 } @article{1304546, abstract = {Social bookmarking services have recently gained popularity among Web users. Whereas numerous studies provide a historical account of tagging systems, the authors use their analysis of a domain-specific social bookmarking service called CiteULike to reflect on two metrics for evaluating tagging behavior: tag growth and tag reuse. They examine the relationship between these two metrics and articulate design implications for enhancing social bookmarking services. The authors also briefly reflect on their own work on developing a social bookmarking service for CiteSeer, an online scholarly digital library for computer science.}, address = {Piscataway, NJ, USA}, author = {Farooq, Umer and Song, Yang and Carroll, John M. and Giles, C. Lee}, doi = {http://dx.doi.org/10.1109/MIC.2007.135}, interhash = {13183e8fc4cbe0944a819afa2d9ff4eb}, intrahash = {5785e8a8064b3d346f8c198c3c860bf6}, issn = {1089-7801}, journal = {IEEE Internet Computing}, number = 6, pages = {29--35}, publisher = {IEEE Educational Activities Department}, title = {Social Bookmarking for Scholarly Digital Libraries}, url = {http://portal.acm.org/citation.cfm?id=1304546&coll=Portal&dl=GUIDE&CFID=46454031&CFTOKEN=27530397}, volume = 11, year = 2007 } @inproceedings{1390423, address = {New York, NY, USA}, author = {Song, Yang and Zhuang, Ziming and Li, Huajing and Zhao, Qiankun and Li, Jia and Lee, Wang-Chien and Giles, C. Lee}, booktitle = {SIGIR '08: Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval}, doi = {http://doi.acm.org/10.1145/1390334.1390423}, interhash = {e6505664e875de06d98a6e787d4367d1}, intrahash = {525a37f6ef3d81a81686b515a148b88b}, isbn = {978-1-60558-164-4}, location = {Singapore, Singapore}, pages = {515--522}, publisher = {ACM}, title = {Real-time automatic tag recommendation}, url = {http://portal.acm.org/citation.cfm?id=1390334.1390423}, year = 2008 } @inproceedings{1458098, address = {New York, NY, USA}, author = {Song, Yang and Zhang, Lu and Giles, C. Lee}, booktitle = {CIKM '08: Proceeding of the 17th ACM conference on Information and knowledge mining}, doi = {http://doi.acm.org/10.1145/1458082.1458098}, interhash = {5c03bc1e658b6d44f053944418bdaec3}, intrahash = {d330a3537b4a14fbd40661424ec8e465}, isbn = {978-1-59593-991-3}, location = {Napa Valley, California, USA}, pages = {93--102}, publisher = {ACM}, title = {A sparse gaussian processes classification framework for fast tag suggestions}, url = {http://portal.acm.org/citation.cfm?id=1458098}, year = 2008 } @inproceedings{1316677, address = {New York, NY, USA}, author = {Farooq, Umer and Kannampallil, Thomas G. and Song, Yang and Ganoe, Craig H. and Carroll, John M. and Giles, Lee}, booktitle = {GROUP '07: Proceedings of the 2007 international ACM conference on Conference on supporting group work}, doi = {http://doi.acm.org/10.1145/1316624.1316677}, interhash = {66928ca91bf0d777b848fe6f7a55de20}, intrahash = {5d0b61727d81aed019ba4297090108ca}, isbn = {978-1-59593-845-9}, location = {Sanibel Island, Florida, USA}, pages = {351--360}, publisher = {ACM}, title = {Evaluating tagging behavior in social bookmarking systems: metrics and design heuristics}, url = {http://portal.acm.org/citation.cfm?id=1316677&coll=Portal&dl=GUIDE&CFID=9767993&CFTOKEN=86305662}, year = 2007 }