In this paper, we present the factorization model PITF (Pairwise Interaction Tensor Factorization) which is a special case of the TD model with linear runtime both for learning and prediction. PITF explicitly models the pairwise interactions between users, items and tags. The model is learned with an adaption of the Bayesian personalized ranking (BPR) criterion which originally has been introduced for item recommendation. Empirically, we show on real world datasets that this model outperforms TD largely in runtime and even can achieve better prediction quality. Besides our lab experiments, PITF has also won the ECML/PKDD Discovery Challenge 2009 for graph-based tag recommendation.}, acmid = {1718498}, address = {New York, NY, USA}, author = {Rendle, Steffen and Schmidt-Thieme, Lars}, booktitle = {Proceedings of the third ACM international conference on Web search and data mining}, doi = {10.1145/1718487.1718498}, interhash = {ce8fbdf2afb954579cdb58104fb683a7}, intrahash = {10fe730b391b08031f3103f9cdbb6e1a}, isbn = {978-1-60558-889-6}, location = {New York, New York, USA}, numpages = {10}, pages = {81--90}, publisher = {ACM}, series = {WSDM '10}, title = {Pairwise interaction tensor factorization for personalized tag recommendation}, url = {http://doi.acm.org/10.1145/1718487.1718498}, year = 2010 } @inproceedings{rendle2010pairwise, abstract = {Tagging plays an important role in many recent websites. Recommender systems can help to suggest a user the tags he might want to use for tagging a specific item. Factorization models based on the Tucker Decomposition (TD) model have been shown to provide high quality tag recommendations outperforming other approaches like PageRank, FolkRank, collaborative filtering, etc. The problem with TD models is the cubic core tensor resulting in a cubic runtime in the factorization dimension for prediction and learning.
In this paper, we present the factorization model PITF (Pairwise Interaction Tensor Factorization) which is a special case of the TD model with linear runtime both for learning and prediction. PITF explicitly models the pairwise interactions between users, items and tags. The model is learned with an adaption of the Bayesian personalized ranking (BPR) criterion which originally has been introduced for item recommendation. Empirically, we show on real world datasets that this model outperforms TD largely in runtime and even can achieve better prediction quality. Besides our lab experiments, PITF has also won the ECML/PKDD Discovery Challenge 2009 for graph-based tag recommendation.}, acmid = {1718498}, address = {New York, NY, USA}, author = {Rendle, Steffen and Schmidt-Thieme, Lars}, booktitle = {Proceedings of the third ACM international conference on Web search and data mining}, doi = {10.1145/1718487.1718498}, interhash = {ce8fbdf2afb954579cdb58104fb683a7}, intrahash = {10fe730b391b08031f3103f9cdbb6e1a}, isbn = {978-1-60558-889-6}, location = {New York, New York, USA}, numpages = {10}, pages = {81--90}, publisher = {ACM}, title = {Pairwise interaction tensor factorization for personalized tag recommendation}, url = {http://doi.acm.org/10.1145/1718487.1718498}, year = 2010 } @incollection{marinho2011social, abstract = {The new generation of Web applications known as (STS) is successfully established and poised for continued growth. STS are open and inherently social; features that have been proven to encourage participation. But while STS bring new opportunities, they revive old problems, such as information overload. Recommender Systems are well known applications for increasing the level of relevant content over the noise that continuously grows as more and more content becomes available online. In STS however, we face new challenges. Users are interested in finding not only content, but also tags and even other users. Moreover, while traditional recommender systems usually operate over 2-way data arrays, STS data is represented as a third-order tensor or a hypergraph with hyperedges denoting (user, resource, tag) triples. In this chapter, we survey the most recent and state-of-the-art work about a whole new generation of recommender systems built to serve STS.We describe (a) novel facets of recommenders for STS, such as user, resource, and tag recommenders, (b) new approaches and algorithms for dealing with the ternary nature of STS data, and (c) recommender systems deployed in real world STS. Moreover, a concise comparison between existing works is presented, through which we identify and point out new research directions.}, address = {New York}, author = {Balby Marinho, Leandro and Nanopoulos, Alexandros and Schmidt-Thieme, Lars and Jäschke, Robert and Hotho, Andreas and Stumme, Gerd and Symeonidis, Panagiotis}, booktitle = {Recommender Systems Handbook}, doi = {10.1007/978-0-387-85820-3_19}, editor = {Ricci, Francesco and Rokach, Lior and Shapira, Bracha and Kantor, Paul B.}, interhash = {2d4afa6f7fb103ccc166c9c5d629cdd1}, intrahash = {708be7b5c269bd3a9d3d2334f858d52d}, isbn = {978-0-387-85820-3}, pages = {615--644}, publisher = {Springer}, title = {Social Tagging Recommender Systems}, url = {http://dx.doi.org/10.1007/978-0-387-85820-3_19}, year = 2011 } @inproceedings{benz2006automatic, abstract = {Bookmarks (or Favorites, Hotlists) are a popular strategy to relocate interesting websites on the WWW by creating a personalized local URL repository. Most current browsers offer a facility to store and manage bookmarks in a hierarchy of folders; though, with growing size, users reportedly have trouble to create and maintain a stable taxonomy. This paper presents a novel collaborative approach to ease bookmark management, especially the “classification?? of new bookmarks into a folder. We propose a methodology to realize the collaborative classification idea of considering how similar users have classified a bookmark. A combination of nearest-neighbour-classifiers is used to derive a recommendation from similar users on where to store a new bookmark. Additionally, a procedure to generate keyword recommendations is proposed to ease the annotation of new bookmarks. A prototype system called CariBo has been implemented as a plugin of the central bookmark server software SiteBar. A case study conducted with real user data supports the validity of the approach.}, address = {Edinburgh, Scotland}, author = {Benz, Dominik and Tso, Karen H. L. and Schmidt-Thieme, Lars}, booktitle = {Proceedings of the 2nd Workshop in Innovations in Web Infrastructure (IWI2) at WWW2006}, file = {benz2006automatic.pdf:benz2006automatic.pdf:PDF}, interhash = {72efbe1b69d12bc3ce35522bc1c83e82}, intrahash = {9d685c05008804a45b72a43586777b3b}, month = may, note = {isbn = {085432853X}}, title = {Automatic Bookmark Classification - A Collaborative Approach}, url = {http://www.kde.cs.uni-kassel.de/pub/pdf/benz2006automatic.pdf}, year = 2006 } @inproceedings{jaeschke2007tag, abstract = {Collaborative tagging systems allow users to assign keywords—so called “tags”—to resources. Tags are used for navigation, finding resources and serendipitous browsing and thus provide an immediate benefit for users. These systems usually include tag recommendation mechanisms easing the process of finding good tags for a resource, but also consolidating the tag vocabulary across users. In practice, however, only very basic recommendation strategies are applied. In this paper we evaluate and compare two recommendation algorithms on largescale real life datasets: an adaptation of user-based collaborative filtering and a graph-based recommender built on top of FolkRank. We show that both provide better results than non-personalized baseline methods. Especially the graph-based recommender outperforms existing methods considerably.}, address = {Berlin, Heidelberg}, author = {Jäschke, Robert and Marinho, Leandro Balby and Hotho, Andreas and Schmidt-Thieme, Lars and Stumme, Gerd}, booktitle = {Knowledge Discovery in Databases: PKDD 2007, 11th European Conference on Principles and Practice of Knowledge Discovery in Databases}, editor = {Kok, Joost N. and Koronacki, Jacek and de Mántaras, Ramon López and Matwin, Stan and Mladenic, Dunja and Skowron, Andrzej}, ee = {http://dx.doi.org/10.1007/978-3-540-74976-9_52}, interhash = {7e212e3bac146d406035adebff248371}, intrahash = {bb8ecec699a2f129322fe334747c6aef}, isbn = {978-3-540-74975-2}, pages = {506-514}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, title = {Tag Recommendations in Folksonomies}, url = {http://dx.doi.org/10.1007/978-3-540-74976-9_52}, volume = 4702, year = 2007 } @article{jaeschke2008tag, abstract = {Collaborative tagging systems allow users to assign keywords - so called "tags" - to resources. Tags are used for navigation, finding resources and serendipitous browsing and thus provide an immediate benefit for users. These systems usually include tag recommendation mechanisms easing the process of finding good tags for a resource, but also consolidating the tag vocabulary across users. In practice, however, only very basic recommendation strategies are applied. In this paper we evaluate and compare several recommendation algorithms on large-scale real life datasets: an adaptation of user-based collaborative filtering, a graph-based recommender built on top of the FolkRank algorithm, and simple methods based on counting tag occurences. We show that both FolkRank and Collaborative Filtering provide better results than non-personalized baseline methods. Moreover, since methods based on counting tag occurrences are computationally cheap, and thus usually preferable for real time scenarios, we discuss simple approaches for improving the performance of such methods. We show, how a simple recommender based on counting tags from users and resources can perform almost as good as the best recommender. }, address = {Amsterdam}, author = {Jäschke, Robert and Marinho, Leandro and Hotho, Andreas and Schmidt-Thieme, Lars and Stumme, Gerd}, doi = {10.3233/AIC-2008-0438}, editor = {Giunchiglia, Enrico}, interhash = {b2f1aba6829affc85d852ea93a8e39f7}, intrahash = {955bcf14f3272ba6eaf3dadbef6c0b10}, issn = {0921-7126}, journal = {AI Communications}, number = 4, pages = {231-247}, publisher = {IOS Press}, title = {Tag Recommendations in Social Bookmarking Systems}, url = {http://dx.doi.org/10.3233/AIC-2008-0438}, vgwort = {63}, volume = 21, year = 2008 } @inproceedings{jaeschke07tagKdml, author = {Jaeschke, Robert and Marinho, Leandro and Hotho, Andreas and Schmidt-Thieme, Lars and Stumme, Gerd}, booktitle = {Workshop Proceedings of Lernen - Wissensentdeckung - Adaptivität (LWA 2007)}, editor = {Hinneburg, Alexander}, interhash = {19e40fd1eb137fab091512656ecc504d}, intrahash = {71bc9f8ae1a53632dc9a2b98b017f152}, isbn = {978-3-86010-907-6}, month = sep, pages = {13-20}, publisher = {Martin-Luther-Universität Halle-Wittenberg}, title = {Tag Recommendations in Folksonomies}, url = {http://www.kde.cs.uni-kassel.de/stumme/papers/2007/jaeschke07tagrecommendationsKDML.pdf}, year = 2007 } @article{benz2007supporting, abstract = {Bookmarks (or favorites, hotlists) are popular strategies to relocate interesting websites on the WWW by creating a personalized URL repository. Most current browsers offer a facility to locally store and manage bookmarks in a hierarchy of folders; though, with growing size, users reportedly have trouble to create and maintain a stable organization structure. This paper presents a novel collaborative approach to ease bookmark management, especially the “classification?? of new bookmarks into a folder. We propose a methodology to realize the collaborative classification idea of considering how similar users have classified a bookmark. A combination of nearest-neighbor-classifiers is used to derive a recommendation from similar users on where to store a new bookmark. A prototype system called CariBo has been implemented as a plugin for the central bookmark server software SiteBar. All findings have been evaluated on a reasonably large scale, real user dataset with promising results, and possible implications for shared and social bookmarking systems are discussed.}, author = {Benz, Dominik and Tso, Karen H. L. and Schmidt-Thieme, Lars}, doi = {http://dx.doi.org/10.1016/j.comnet.2007.06.014}, file = {benz2007supporting.pdf:benz2007supporting.pdf:PDF}, interhash = {181404c6a55baf6fe5db8448ac0d5bf0}, intrahash = {ec4387ab22950b38a0c152d4f33b0987}, journal = {Special Issue of the Computer Networks journal on Innovations in Web Communications Infrastructure}, number = 16, pages = {4574--4585}, title = {Supporting Collaborative Hierarchical Classification: Bookmarks as an Example}, url = {http://www.kde.cs.uni-kassel.de/pub/pdf/benz2007supporting.pdf}, volume = 51, year = 2007 } @inproceedings{Rendle:2010:FPM:1772690.1772773, abstract = {Recommender systems are an important component of many websites. Two of the most popular approaches are based on matrix factorization (MF) and Markov chains (MC). MF methods learn the general taste of a user by factorizing the matrix over observed user-item preferences. On the other hand, MC methods model sequential behavior by learning a transition graph over items that is used to predict the next action based on the recent actions of a user. In this paper, we present a method bringing both approaches together. Our method is based on personalized transition graphs over underlying Markov chains. That means for each user an own transition matrix is learned - thus in total the method uses a transition cube. As the observations for estimating the transitions are usually very limited, our method factorizes the transition cube with a pairwise interaction model which is a special case of the Tucker Decomposition. We show that our factorized personalized MC (FPMC) model subsumes both a common Markov chain and the normal matrix factorization model. For learning the model parameters, we introduce an adaption of the Bayesian Personalized Ranking (BPR) framework for sequential basket data. Empirically, we show that our FPMC model outperforms both the common matrix factorization and the unpersonalized MC model both learned with and without factorization.}, acmid = {1772773}, address = {New York, NY, USA}, author = {Rendle, Steffen and Freudenthaler, Christoph and Lars, Schmidt-Thieme}, booktitle = {Proceedings of the 19th international conference on World wide web}, doi = {10.1145/1772690.1772773}, interhash = {578c4d7bb47620c5b943d9533ae2ae5f}, intrahash = {a806788d8f8f9259624a853551e40c30}, isbn = {978-1-60558-799-8}, location = {Raleigh, North Carolina, USA}, numpages = {10}, pages = {811--820}, publisher = {ACM}, series = {WWW '10}, title = {Factorizing personalized Markov chains for next-basket recommendation}, url = {http://doi.acm.org/10.1145/1772690.1772773}, year = 2010 } @incollection{marinho2011social, abstract = {The new generation of Web applications known as (STS) is successfully established and poised for continued growth. STS are open and inherently social; features that have been proven to encourage participation. But while STS bring new opportunities, they revive old problems, such as information overload. Recommender Systems are well known applications for increasing the level of relevant content over the noise that continuously grows as more and more content becomes available online. In STS however, we face new challenges. Users are interested in finding not only content, but also tags and even other users. Moreover, while traditional recommender systems usually operate over 2-way data arrays, STS data is represented as a third-order tensor or a hypergraph with hyperedges denoting (user, resource, tag) triples. In this chapter, we survey the most recent and state-of-the-art work about a whole new generation of recommender systems built to serve STS.We describe (a) novel facets of recommenders for STS, such as user, resource, and tag recommenders, (b) new approaches and algorithms for dealing with the ternary nature of STS data, and (c) recommender systems deployed in real world STS. Moreover, a concise comparison between existing works is presented, through which we identify and point out new research directions.}, address = {New York}, author = {Balby Marinho, Leandro and Nanopoulos, Alexandros and Schmidt-Thieme, Lars and Jäschke, Robert and Hotho, Andreas and Stumme, Gerd and Symeonidis, Panagiotis}, booktitle = {Recommender Systems Handbook}, doi = {10.1007/978-0-387-85820-3_19}, editor = {Ricci, Francesco and Rokach, Lior and Shapira, Bracha and Kantor, Paul B.}, interhash = {2d4afa6f7fb103ccc166c9c5d629cdd1}, intrahash = {708be7b5c269bd3a9d3d2334f858d52d}, isbn = {978-0-387-85820-3}, pages = {615--644}, publisher = {Springer}, title = {Social Tagging Recommender Systems}, url = {http://dx.doi.org/10.1007/978-0-387-85820-3_19}, vgwort = {50}, year = 2011 } @inproceedings{haase2005usagedriven, abstract = {Large information repositories as digital libraries, online shops, etc. rely on a taxonomy of the objects under consideration to structure the vast contents and facilitate browsing and searching (e.g., ACM topic classification for computer science literature, Amazon product taxonomy, etc.). As in heterogeneous communities users typically will use different parts of such an ontology with varying intensity, customization and personalization of the ontologies is desirable. In this paper we adapt a collaborative filtering recommender system to assist users in the management and evolution of their personal ontology by providing detailed suggestions of ontology changes. Such a system has been implemented in the context of Bibster, a peer-to-peer based personal bibliography management tool. Finally, we report on an in-situ experiment with the Bibster community that shows the performance improvements over non-personalized recommendations.}, address = {Las Vegas, Nevada USA}, author = {Haase, Peter and Hotho, Andreas and Schmidt-Thieme, Lars and Sure, York}, booktitle = {Proceedings of the 3rd International Conference on Universal A ccess in Human-Computer Interaction (UAHCI)}, file = {haase2005usagedriven.pdf:haase2005usagedriven.pdf:PDF}, groups = {public}, interhash = {e5681c379cdfe126e44d034dac3fddad}, intrahash = {05b1daa45aedb5c1f63683f961f17a9e}, lastdatemodified = {2006-07-06}, lastname = {Haase}, month = {22-27 July}, own = {notown}, pdf = {Haase05.pdf}, read = {notread}, timestamp = {2007-05-25 16:05:53}, title = {Usage-driven Evolution of Personal Ontologies}, username = {dbenz}, year = 2005 } @inproceedings{haase2005collaborative, abstract = {Large information repositories as digital libraries, online shops, etc. rely on a taxonomy of the objects under consideration to structure the vast contents and facilitate browsing and searching (e.g., ACM topic classification for computer science literature, Amazon product taxonomy, etc.). As in heterogenous communities users typically will use different parts of such an ontology with varying intensity, customization and personalization of the ontologies is desirable. Of particular interest for supporting users during the personalization are collaborative filtering systems which can produce personal recommendations by computing the similarity between own preferences and the one of other people. In this paper we adapt a collaborative filtering recommender system to assist users in the management and evolution of their personal ontology by providing detailed suggestions of ontology changes. Such a system has been implemented in the context of Bibster, a peer-to-peer based personal bibliography management tool. Finally, we report on an experiment with the Bibster community that shows the performance improvements over non-personalized recommendations.}, author = {Haase, Peter and Hotho, Andreas and Schmidt-Thieme, Lars and Sure, York}, booktitle = {ESWC}, crossref = {conf/esws/2005}, date = {2005-05-24}, editor = {Gómez-Pérez, Asunción and Euzenat, Jérôme}, ee = {http://dx.doi.org/10.1007/11431053_33}, file = {haase2005collaborative.pdf:haase2005collaborative.pdf:PDF}, groups = {public}, interhash = {c9ba81293a1b27f1c9bdf38a3beec060}, intrahash = {1a8829cde1cb26241a48901e28a953d2}, isbn = {3-540-26124-9}, pages = {486-499}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, timestamp = {2009-11-10 11:30:42}, title = {Collaborative and Usage-Driven Evolution of Personal Ontologies.}, url = {http://www.aifb.uni-karlsruhe.de/WBS/pha/publications/collaborative05eswc.pdf}, username = {dbenz}, volume = 3532, year = 2005 } @inproceedings{marinho2008folksonomybased, abstract = {The growing popularity of social tagging systems promises to alleviate the knowledge bottleneck that slows down the full materialization of the SemanticWeb since these systems allow ordinary users to create and share knowledge in a simple, cheap, and scalable representation, usually known as folksonomy. However, for the sake of knowledge workflow, one needs to find a compromise between the uncontrolled nature of folksonomies and the controlled and more systematic vocabulary of domain experts. In this paper we propose to address this concern by devising a method that automatically enriches a folksonomy with domain expert knowledge and by introducing a novel algorithm based on frequent itemset mining techniques to efficiently learn an ontology over the enriched folksonomy. In order to quantitatively assess our method, we propose a new benchmark for task-based ontology evaluation where the quality of the ontologies is measured based on how helpful they are for the task of personalized information finding. We conduct experiments on real data and empirically show the effectiveness of our approach.}, author = {Marinho, Leandro Balby and Buza, Krisztian and Schmidt-Thieme, Lars}, booktitle = {International Semantic Web Conference}, crossref = {conf/semweb/2008}, date = {2008-10-24}, editor = {Sheth, Amit P. and Staab, Steffen and Dean, Mike and Paolucci, Massimo and Maynard, Diana and Finin, Timothy W. and Thirunarayan, Krishnaprasad}, ee = {http://dx.doi.org/10.1007/978-3-540-88564-1_17}, file = {marinho2008folksonomybased.pdf:marinho2008folksonomybased.pdf:PDF}, groups = {public}, interhash = {d295e7d4615500c670e70ad240fada29}, intrahash = {cfa4c4520d4cf02e03dd3b84bb5c9578}, isbn = {978-3-540-88563-4}, pages = {261-276}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, timestamp = {2010-03-30 16:14:58}, title = {Folksonomy-Based Collabulary Learning.}, url = {http://dblp.uni-trier.de/db/conf/semweb/iswc2008.html#MarinhoBS08}, username = {dbenz}, volume = 5318, year = 2008 } @inproceedings{benz2006automatic, abstract = {Bookmarks (or Favorites, Hotlists) are a popular strategy to relocate interesting websites on the WWW by creating a personalized local URL repository. Most current browsers offer a facility to store and manage bookmarks in a hierarchy of folders; though, with growing size, users reportedly have trouble to create and maintain a stable taxonomy. This paper presents a novel collaborative approach to ease bookmark management, especially the “classification�? of new bookmarks into a folder. We propose a methodology to realize the collaborative classification idea of considering how similar users have classified a bookmark. A combination of nearest-neighbour-classifiers is used to derive a recommendation from similar users on where to store a new bookmark. Additionally, a procedure to generate keyword recommendations is proposed to ease the annotation of new bookmarks. A prototype system called CariBo has been implemented as a plugin of the central bookmark server software SiteBar. A case study conducted with real user data supports the validity of the approach.}, address = {Edinburgh, Scotland}, author = {Benz, Dominik and Tso, Karen H. L. and Schmidt-Thieme, Lars}, booktitle = {Proceedings of the 2nd Workshop in Innovations in Web Infrastructure (IWI2) at WWW2006}, file = {benz2006automatic.pdf:benz2006automatic.pdf:PDF}, interhash = {72efbe1b69d12bc3ce35522bc1c83e82}, intrahash = {9d685c05008804a45b72a43586777b3b}, month = May, note = {isbn = {085432853X}}, title = {Automatic Bookmark Classification - A Collaborative Approach}, url = {http://www.kde.cs.uni-kassel.de/pub/pdf/benz2006automatic.pdf}, year = 2006 } @article{benz2007supporting, abstract = {Bookmarks (or favorites, hotlists) are popular strategies to relocate interesting websites on the WWW by creating a personalized URL repository. Most current browsers offer a facility to locally store and manage bookmarks in a hierarchy of folders; though, with growing size, users reportedly have trouble to create and maintain a stable organization structure. This paper presents a novel collaborative approach to ease bookmark management, especially the “classification�? of new bookmarks into a folder. We propose a methodology to realize the collaborative classification idea of considering how similar users have classified a bookmark. A combination of nearest-neighbor-classifiers is used to derive a recommendation from similar users on where to store a new bookmark. A prototype system called CariBo has been implemented as a plugin for the central bookmark server software SiteBar. All findings have been evaluated on a reasonably large scale, real user dataset with promising results, and possible implications for shared and social bookmarking systems are discussed.}, author = {Benz, Dominik and Tso, Karen H. L. and Schmidt-Thieme, Lars}, doi = {http://dx.doi.org/10.1016/j.comnet.2007.06.014}, file = {benz2007supporting.pdf:benz2007supporting.pdf:PDF}, interhash = {181404c6a55baf6fe5db8448ac0d5bf0}, intrahash = {ec4387ab22950b38a0c152d4f33b0987}, journal = {Special Issue of the Computer Networks journal on Innovations in Web Communications Infrastructure}, number = 16, pages = {4574--4585}, title = {Supporting Collaborative Hierarchical Classification: Bookmarks as an Example}, url = {http://www.kde.cs.uni-kassel.de/pub/pdf/benz2007supporting.pdf}, volume = 51, year = 2007 } @article{jaeschke2008tag, abstract = {Collaborative tagging systems allow users to assign keywords - so called "tags" - to resources. Tags are used for navigation, finding resources and serendipitous browsing and thus provide an immediate benefit for users. These systems usually include tag recommendation mechanisms easing the process of finding good tags for a resource, but also consolidating the tag vocabulary across users. In practice, however, only very basic recommendation strategies are applied. In this paper we evaluate and compare several recommendation algorithms on large-scale real life datasets: an adaptation of user-based collaborative filtering, a graph-based recommender built on top of the FolkRank algorithm, and simple methods based on counting tag occurences. We show that both FolkRank and Collaborative Filtering provide better results than non-personalized baseline methods. Moreover, since methods based on counting tag occurrences are computationally cheap, and thus usually preferable for real time scenarios, we discuss simple approaches for improving the performance of such methods. We show, how a simple recommender based on counting tags from users and resources can perform almost as good as the best recommender. }, address = {Amsterdam}, author = {Jäschke, Robert and Marinho, Leandro and Hotho, Andreas and Schmidt-Thieme, Lars and Stumme, Gerd}, doi = {10.3233/AIC-2008-0438}, editor = {Giunchiglia, Enrico}, interhash = {b2f1aba6829affc85d852ea93a8e39f7}, intrahash = {955bcf14f3272ba6eaf3dadbef6c0b10}, issn = {0921-7126}, journal = {AI Communications}, month = dec, number = 4, pages = {231--247}, publisher = {IOS Press}, title = {Tag Recommendations in Social Bookmarking Systems}, url = {http://www.kde.cs.uni-kassel.de/pub/pdf/jaeschke2008tag.pdf}, vgwort = {63}, volume = 21, year = 2008 } @incollection{reference/rsh/MarinhoNSJHSS11, author = {Marinho, Leandro Balby and Nanopoulos, Alexandros and Schmidt-Thieme, Lars and Jäschke, Robert and Hotho, Andreas and Stumme, Gerd and Symeonidis, Panagiotis}, booktitle = {Recommender Systems Handbook}, crossref = {reference/rsh/2011}, editor = {Ricci, Francesco and Rokach, Lior and Shapira, Bracha and Kantor, Paul B.}, ee = {http://dx.doi.org/10.1007/978-0-387-85820-3_19}, interhash = {2d4afa6f7fb103ccc166c9c5d629cdd1}, intrahash = {8a520671b6ced7c4b81b1cd18274e0ee}, isbn = {978-0-387-85819-7}, pages = {615-644}, publisher = {Springer}, title = {Social Tagging Recommender Systems.}, url = {http://dblp.uni-trier.de/db/reference/rsh/rsh2011.html#MarinhoNSJHSS11}, year = 2011 } @inproceedings{marinho:ecml2009, abstract = {This paper describes our approach to the ECML/PKDD Discovery Challenge 2009. Our approach is a pure statistical model taking no content information into account. It tries to find latent interactions between users, items and tags by factorizing the observed tagging data. The factorization model is learned by the Bayesian Personal Ranking method (BPR) which is inspired by a Bayesian analysis of personalized ranking with missing data. To prevent overfitting, we ensemble the models over several iterations and hyperparameters. Finally, we enhance the top-n lists by estimating how many tags to recommend.}, address = {Bled, Slovenia}, author = {Rendle, Steffen and Schmidt-Thieme, Lars}, booktitle = {ECML PKDD Discovery Challenge 2009 (DC09)}, editor = {Eisterlehner, Folke and Hotho, Andreas and Jäschke, Robert}, interhash = {8485850cde1a6b61971cac27fa867845}, intrahash = {ceed045a84e121fa37384f797306d30f}, issn = {1613-0073}, month = {September}, pages = {235--242}, publisher = {CEUR Workshop Proceedings}, title = {Factor Models for Tag Recommendation in BibSonomy}, url = {http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-497/}, volume = 497, year = 2009 }