Recommendation algorithms and multi-class classifiers can support users of social bookmarking systems in assigning tags to their bookmarks. Content based recommenders are the usual approach for facing the cold start problem, i.e., when a bookmark is uploaded for the first time and no information from other users can be exploited. In this paper, we evaluate several recommendation algorithms in a cold-start scenario on a large real-world dataset.

@inproceedings{illig2009comparison, author = {Illig, Jens and Hotho, Andreas and Jäschke, Robert and Stumme, Gerd}, title = {A Comparison of Content-Based Tag Recommendations in Folksonomy Systems}, editor = {Wolff, Karl Erich and Palchunov, Dmitry E. and Zagoruiko, Nikolay G. and Andelfinger, Urs}, booktitle = {Knowledge Processing and Data Analysis}, series = {Lecture Notes in Computer Science}, publisher = {Springer}, address = {Berlin/Heidelberg}, year = {2011}, volume = {6581}, pages = {136--149}, url = {http://dx.doi.org/10.1007/978-3-642-22140-8_9}, doi = {10.1007/978-3-642-22140-8_9}, isbn = {978-3-642-22139-2}, keywords = {2011, folksonomy, myown, recommender, tag}, abstract = {Recommendation algorithms and multi-class classifiers can support users of social bookmarking systems in assigning tags to their bookmarks. Content based recommenders are the usual approach for facing the cold start problem, i.e., when a bookmark is uploaded for the first time and no information from other users can be exploited. In this paper, we evaluate several recommendation algorithms in a cold-start scenario on a large real-world dataset. } }

%0 = inproceedings %A = Illig, Jens and Hotho, Andreas and Jäschke, Robert and Stumme, Gerd %B = Knowledge Processing and Data Analysis %C = Berlin/Heidelberg %D = 2011 %I = Springer %T = A Comparison of Content-Based Tag Recommendations in Folksonomy Systems %U = http://dx.doi.org/10.1007/978-3-642-22140-8_9

One of the most noticeable innovation that emerged with the advent of the Web 2.0 and the focal point of this thesis are collaborative tagging systems. They allow users to annotate arbitrary resources with freely chosen keywords, so called tags. The tags are used for navigation, finding resources, and serendipitous browsing and thus provide an immediate benefit for the user. By now, several systems for tagging photos, web links, publication references, videos, etc. have attracted millions of users which in turn annotated countless resources. Tagging gained so much popularity that it spread into other applications like web browsers, software packet managers, and even file systems. Therefore, the relevance of the methods presented in this thesis goes beyond the Web 2.0. The conceptual structure underlying collaborative tagging systems is called folksonomy. It can be represented as a tripartite hypergraph with user, tag, and resource nodes. Each edge of the graph expresses the fact that a user annotated a resource with a tag. This social network constitutes a lightweight conceptual structure that is not formalized, but rather implicit and thus needs to be extracted with knowledge discovery methods. In this thesis a new data mining task – the mining of all frequent tri-concepts – is presented, together with an efficient algorithm for discovering such implicit shared conceptualizations. Our approach extends the data mining task of discovering all closed itemsets to three-dimensional data structures to allow for mining folksonomies. Extending the theory of triadic Formal Concept Analysis, we provide a formal definition of the problem, and present an efficient algorithm for its solution. We show the applicability of our approach on three large real-world examples and thereby perform a conceptual clustering of two collaborative tagging systems. Finally, we introduce neighborhoods of triadic concepts as basis for a lightweight visualization of tri-lattices. The social bookmark and publication sharing system BibSonomy, which is currently among the three most popular systems of its kind, has been developed by our research group. Besides being a useful tool for many scientists, it provides interested researchers a basis for the evaluation and integration of their knowledge discovery methods. This thesis introduces BibSonomy as an exemplary collaborative tagging system and gives an overview of its architecture and some of its features. Furthermore, BibSonomy is used as foundation for evaluating and integrating some of the discussed approaches. Collaborative tagging 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 this thesis we evaluate and compare several recommendation algorithms on large-scale real-world 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 co-occurences. We show that both FolkRank and Collaborative Filtering provide better results than non-personalized baseline methods. Moreover, since methods based on counting tag co-occurrences are computationally cheap, and thus usually preferable for real time scenarios, we discuss simple approaches for improving the performance of such methods. We demonstrate how a simple recommender based on counting tags from users and resources can perform almost as good as the best recommender. Furthermore, we show how to integrate recommendation methods into a real tagging system, record and evaluate their performance by describing the tag recommendation framework we developed for BibSonomy. With the intention to develop, test, and evaluate recommendation algorithms and supporting cooperation with researchers, we designed the framework to be easily extensible, open for a variety of methods, and usable independent from BibSonomy. We also present an evaluation of the framework which demonstrates its power. The folksonomy graph shows specific structural properties that explain its growth and the possibility of serendipitous exploration. Clicklogs of web search engines can be represented as a folksonomy in which queries are descriptions of clicked URLs. The resulting network structure, which we will term logsonomy is very similar to the one of folksonomies. In order to find out about its properties, we analyze the topological characteristics of the tripartite hypergraph of queries, users and bookmarks on a large folksonomy snapshot and on query logs of two large search engines. We find that all of the three datasets exhibit similar structural properties and thus conclude that the clicking behaviour of search engine users based on the displayed search results and the tagging behaviour of collaborative tagging users is driven by similar dynamics. In this thesis we further transfer the folksonomy paradigm to the Social Semantic Desktop – a new model of computer desktop that uses Semantic Web technologies to better link information items. There we apply community support methods to the folksonomy found in the network of social semantic desktops. Thus, we connect knowledge discovery for folksonomies with semantic technologies. Alltogether, the research in this thesis is centered around collaborative tagging systems and their underlying datastructure – folksonomies – and thereby paves the way for the further dissemination of this successful knowledge management paradigm.

@book{jaeschke2011formal, author = {Jäschke, Robert}, title = {Formal Concept Analysis and Tag Recommendations in Collaborative Tagging Systems}, series = {Dissertationen zur Künstlichen Intelligenz}, publisher = {Akademische Verlagsgesellschaft AKA}, address = {Heidelberg, Germany}, year = {2011}, volume = {332}, url = {http://www.aka-verlag.com/de/detail?ean=978-3-89838-332-5}, isbn = {978-3-89838-332-5}, keywords = {2011, analysis, collaborative, concept, fca, folksonomy, formal, myown, recommender, tag, tagging}, abstract = {One of the most noticeable innovation that emerged with the advent of the Web 2.0 and the focal point of this thesis are collaborative tagging systems. They allow users to annotate arbitrary resources with freely chosen keywords, so called tags. The tags are used for navigation, finding resources, and serendipitous browsing and thus provide an immediate benefit for the user. By now, several systems for tagging photos, web links, publication references, videos, etc. have attracted millions of users which in turn annotated countless resources. Tagging gained so much popularity that it spread into other applications like web browsers, software packet managers, and even file systems. Therefore, the relevance of the methods presented in this thesis goes beyond the Web 2.0. The conceptual structure underlying collaborative tagging systems is called folksonomy. It can be represented as a tripartite hypergraph with user, tag, and resource nodes. Each edge of the graph expresses the fact that a user annotated a resource with a tag. This social network constitutes a lightweight conceptual structure that is not formalized, but rather implicit and thus needs to be extracted with knowledge discovery methods. In this thesis a new data mining task – the mining of all frequent tri-concepts – is presented, together with an efficient algorithm for discovering such implicit shared conceptualizations. Our approach extends the data mining task of discovering all closed itemsets to three-dimensional data structures to allow for mining folksonomies. Extending the theory of triadic Formal Concept Analysis, we provide a formal definition of the problem, and present an efficient algorithm for its solution. We show the applicability of our approach on three large real-world examples and thereby perform a conceptual clustering of two collaborative tagging systems. Finally, we introduce neighborhoods of triadic concepts as basis for a lightweight visualization of tri-lattices. The social bookmark and publication sharing system BibSonomy, which is currently among the three most popular systems of its kind, has been developed by our research group. Besides being a useful tool for many scientists, it provides interested researchers a basis for the evaluation and integration of their knowledge discovery methods. This thesis introduces BibSonomy as an exemplary collaborative tagging system and gives an overview of its architecture and some of its features. Furthermore, BibSonomy is used as foundation for evaluating and integrating some of the discussed approaches. Collaborative tagging 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 this thesis we evaluate and compare several recommendation algorithms on large-scale real-world 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 co-occurences. We show that both FolkRank and Collaborative Filtering provide better results than non-personalized baseline methods. Moreover, since methods based on counting tag co-occurrences are computationally cheap, and thus usually preferable for real time scenarios, we discuss simple approaches for improving the performance of such methods. We demonstrate how a simple recommender based on counting tags from users and resources can perform almost as good as the best recommender. Furthermore, we show how to integrate recommendation methods into a real tagging system, record and evaluate their performance by describing the tag recommendation framework we developed for BibSonomy. With the intention to develop, test, and evaluate recommendation algorithms and supporting cooperation with researchers, we designed the framework to be easily extensible, open for a variety of methods, and usable independent from BibSonomy. We also present an evaluation of the framework which demonstrates its power. The folksonomy graph shows specific structural properties that explain its growth and the possibility of serendipitous exploration. Clicklogs of web search engines can be represented as a folksonomy in which queries are descriptions of clicked URLs. The resulting network structure, which we will term logsonomy is very similar to the one of folksonomies. In order to find out about its properties, we analyze the topological characteristics of the tripartite hypergraph of queries, users and bookmarks on a large folksonomy snapshot and on query logs of two large search engines. We find that all of the three datasets exhibit similar structural properties and thus conclude that the clicking behaviour of search engine users based on the displayed search results and the tagging behaviour of collaborative tagging users is driven by similar dynamics. In this thesis we further transfer the folksonomy paradigm to the Social Semantic Desktop – a new model of computer desktop that uses Semantic Web technologies to better link information items. There we apply community support methods to the folksonomy found in the network of social semantic desktops. Thus, we connect knowledge discovery for folksonomies with semantic technologies. Alltogether, the research in this thesis is centered around collaborative tagging systems and their underlying datastructure – folksonomies – and thereby paves the way for the further dissemination of this successful knowledge management paradigm. } }

%0 = book %A = Jäschke, Robert %C = Heidelberg, Germany %D = 2011 %I = Akademische Verlagsgesellschaft AKA %T = Formal Concept Analysis and Tag Recommendations in Collaborative Tagging Systems %U = http://www.aka-verlag.com/de/detail?ean=978-3-89838-332-5

In the past decade, Social Tagging Systems have attracted increasing attention from both physical and computer science communities. Besides the underlying structure and dynamics of tagging systems, many efforts have been addressed to unify tagging information to reveal user behaviors and preferences, extract the latent semantic relations among items, make recommendations, and so on. Specifically, this article summarizes recent progress about tag-aware recommender systems, emphasizing on the contributions from three mainstream perspectives and approaches: network-based methods, tensor-based methods, and the topic-based methods. Finally, we outline some other tag-related studies and future challenges of tag-aware recommendation algorithms.

@article{zhang2011tagaware, author = {Zhang, Zi-Ke and Zhou, Tao and Zhang, Yi-Cheng}, title = {Tag-Aware Recommender Systems: A State-of-the-Art Survey}, journal = {Journal of Computer Science and Technology}, publisher = {Springer Boston}, year = {2011}, volume = {26}, number = {5}, pages = {767--777}, url = {http://dx.doi.org/10.1007/s11390-011-0176-1}, doi = {10.1007/s11390-011-0176-1}, issn = {1000-9000}, keywords = {recommender, survey, tag, tagging}, abstract = {In the past decade, Social Tagging Systems have attracted increasing attention from both physical and computer science communities. Besides the underlying structure and dynamics of tagging systems, many efforts have been addressed to unify tagging information to reveal user behaviors and preferences, extract the latent semantic relations among items, make recommendations, and so on. Specifically, this article summarizes recent progress about tag-aware recommender systems, emphasizing on the contributions from three mainstream perspectives and approaches: network-based methods, tensor-based methods, and the topic-based methods. Finally, we outline some other tag-related studies and future challenges of tag-aware recommendation algorithms.} }

%0 = article %A = Zhang, Zi-Ke and Zhou, Tao and Zhang, Yi-Cheng %D = 2011 %I = Springer Boston %T = Tag-Aware Recommender Systems: A State-of-the-Art Survey %U = http://dx.doi.org/10.1007/s11390-011-0176-1

The explosion of collaborative platforms we are recently witnessing, such as social networks, or video and photo sharing sites, radically changed the Web dynamics and the way people use and organize information. The use of tags, keywords freely chosen by users for annotating resources, offers a new way for organizing and retrieving web resources that closely reflects the users' mental model and also allows the use of evolving vocabularies. However, since tags are handled in a purely syntactical way, the annotations provided by users generate a very sparse and noisy tag space that limits the effectiveness of tag-based approaches for complex tasks. Consequently, systems called tag recommenders recently emerged, with the purpose of speeding up the so-called tag convergence, providing users with the most suitable tags for the resource to be annotated. This paper presents a tag recommender system called STaR (Social Tag Recommender), which extends the social approach presented in a previous work [14] with a content-based approach able to extract tags directly from the textual content of HTML pages. Results of experiments carried out on a large dataset gathered from Bibsonomy, show that the use of content-based techniques improves the predictive accuracy of the tag recommender.

@inproceedings{musto2010combining, author = {Musto, Cataldo and Narducci, Fedelucio and Lops, Pasquale and de Gemmis, Marco}, title = {Combining Collaborative and Content-Based Techniques for Tag Recommendation.}, editor = {Buccafurri, Francesco and Semeraro, Giovanni}, booktitle = {E-Commerce and Web Technologies}, series = {Lecture Notes in Business Information Processing}, publisher = {Springer}, address = {Berlin/Heidelberg}, year = {2010}, volume = {61}, pages = {13--23}, url = {http://dx.doi.org/10.1007/978-3-642-15208-5_2}, doi = {10.1007/978-3-642-15208-5_2}, isbn = {978-3-642-15207-8}, keywords = {collaborative, content, recommender, tag, tagging}, abstract = {The explosion of collaborative platforms we are recently witnessing, such as social networks, or video and photo sharing sites, radically changed the Web dynamics and the way people use and organize information. The use of tags, keywords freely chosen by users for annotating resources, offers a new way for organizing and retrieving web resources that closely reflects the users' mental model and also allows the use of evolving vocabularies. However, since tags are handled in a purely syntactical way, the annotations provided by users generate a very sparse and noisy tag space that limits the effectiveness of tag-based approaches for complex tasks. Consequently, systems called tag recommenders recently emerged, with the purpose of speeding up the so-called tag convergence, providing users with the most suitable tags for the resource to be annotated. This paper presents a tag recommender system called STaR (Social Tag Recommender), which extends the social approach presented in a previous work [14] with a content-based approach able to extract tags directly from the textual content of HTML pages. Results of experiments carried out on a large dataset gathered from Bibsonomy, show that the use of content-based techniques improves the predictive accuracy of the tag recommender. } }

%0 = inproceedings %A = Musto, Cataldo and Narducci, Fedelucio and Lops, Pasquale and de Gemmis, Marco %B = E-Commerce and Web Technologies %C = Berlin/Heidelberg %D = 2010 %I = Springer %T = Combining Collaborative and Content-Based Techniques for Tag Recommendation. %U = http://dx.doi.org/10.1007/978-3-642-15208-5_2

In this paper we address the task of recommending additional tags to partially annotated media objects, in our case images. We propose an extendable framework that can recommend tags using a combination of different personalised and collective contexts. We combine information from four contexts: (1) all the photos in the system, (2) a user's own photos, (3) the photos of a user's social contacts, and (4) the photos posted in the groups of which a user is a member. Variants of methods (1) and (2) have been proposed in previous work, but the use of (3) and (4) is novel.</p> <p>For each of the contexts we use the same probabilistic model and Borda Count based aggregation approach to generate recommendations from different contexts into a unified ranking of recommended tags. We evaluate our system using a large set of real-world data from Flickr. We show that by using personalised contexts we can significantly improve tag recommendation compared to using collective knowledge alone. We also analyse our experimental results to explore the capabilities of our system with respect to a user's social behaviour.

@inproceedings{rae2010improving, author = {Rae, Adam and Sigurbjörnsson, Börkur and van Zwol, Roelof}, title = {Improving tag recommendation using social networks}, booktitle = {Adaptivity, Personalization and Fusion of Heterogeneous Information}, series = {RIAO '10}, publisher = {Le Centre De Hautes Etudes Internationales d'Informatique Documentaire}, address = {Paris, France}, year = {2010}, pages = {92--99}, url = {http://portal.acm.org/citation.cfm?id=1937055.1937077}, keywords = {collaborative, recommender, social, tag, tagging}, abstract = {In this paper we address the task of recommending additional tags to partially annotated media objects, in our case images. We propose an extendable framework that can recommend tags using a combination of different personalised and collective contexts. We combine information from four contexts: (1) all the photos in the system, (2) a user's own photos, (3) the photos of a user's social contacts, and (4) the photos posted in the groups of which a user is a member. Variants of methods (1) and (2) have been proposed in previous work, but the use of (3) and (4) is novel.</p> <p>For each of the contexts we use the same probabilistic model and Borda Count based aggregation approach to generate recommendations from different contexts into a unified ranking of recommended tags. We evaluate our system using a large set of real-world data from Flickr. We show that by using personalised contexts we can significantly improve tag recommendation compared to using collective knowledge alone. We also analyse our experimental results to explore the capabilities of our system with respect to a user's social behaviour.} }

%0 = inproceedings %A = Rae, Adam and Sigurbjörnsson, Börkur and van Zwol, Roelof %B = Adaptivity, Personalization and Fusion of Heterogeneous Information %C = Paris, France %D = 2010 %I = Le Centre De Hautes Etudes Internationales d'Informatique Documentaire %T = Improving tag recommendation using social networks %U = http://portal.acm.org/citation.cfm?id=1937055.1937077

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.</p> <p>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.

@inproceedings{rendle2010pairwise, author = {Rendle, Steffen and Schmidt-Thieme, Lars}, title = {Pairwise interaction tensor factorization for personalized tag recommendation}, booktitle = {Proceedings of the third ACM international conference on Web search and data mining}, publisher = {ACM}, address = {New York, NY, USA}, year = {2010}, pages = {81--90}, url = {http://doi.acm.org/10.1145/1718487.1718498}, doi = {10.1145/1718487.1718498}, isbn = {978-1-60558-889-6}, keywords = {collaborative, factorization, folksonomy, personalization, recommender, tag, tagging, tensor}, 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.</p> <p>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.} }

%0 = inproceedings %A = Rendle, Steffen and Schmidt-Thieme, Lars %B = Proceedings of the third ACM international conference on Web search and data mining %C = New York, NY, USA %D = 2010 %I = ACM %T = Pairwise interaction tensor factorization for personalized tag recommendation %U = http://doi.acm.org/10.1145/1718487.1718498

Most tagging systems support the user in the tag selection process by providing tag suggestions, or recommendations, based on a popularity measurement of tags other users provided when tagging the same resource. The majority of theories and mathematical models of tagging found in the literature assume that the emergence of power laws in tagging systems is mainly driven by the imitation behavior of users when observing tag suggestions provided by the user interface of the tagging system. We present experimental results that show that the power law distribution forms regardless of whether or not tag suggestions are presented to the users.

@inproceedings{bollen2009suggestions, author = {Bollen, Dirk and Halpin, Harry}, title = {The Role of Tag Suggestions in Folksonomies}, booktitle = {HT '09: Proceedings of the Twentieth ACM Conference on Hypertext and Hypermedia}, publisher = {ACM}, address = {New York, NY, USA}, year = {2009}, keywords = {folksonomy, ht09, impact, recommender, tag}, abstract = {Most tagging systems support the user in the tag selection process by providing tag suggestions, or recommendations, based on a popularity measurement of tags other users provided when tagging the same resource. The majority of theories and mathematical models of tagging found in the literature assume that the emergence of power laws in tagging systems is mainly driven by the imitation behavior of users when observing tag suggestions provided by the user interface of the tagging system. We present experimental results that show that the power law distribution forms regardless of whether or not tag suggestions are presented to the users.} }

%0 = inproceedings %A = Bollen, Dirk and Halpin, Harry %B = HT '09: Proceedings of the Twentieth ACM Conference on Hypertext and Hypermedia %C = New York, NY, USA %D = 2009 %I = ACM %T = The Role of Tag Suggestions in Folksonomies

@inproceedings{cao2009social, author = {Cao, Hao and Xie, Maoqiang and Xue, Lian and Liu, Chunhua and Teng, Fei and Huang, Yalou}, title = {Social Tag Prediction Base on Supervised Ranking Model}, editor = {Eisterlehner, Folke and Hotho, Andreas and Jäschke, Robert}, series = {CEUR-WS.org}, year = {2009}, volume = {497}, pages = {35--48}, keywords = {dc09, recommender, tag} }

%0 = inproceedings %A = Cao, Hao and Xie, Maoqiang and Xue, Lian and Liu, Chunhua and Teng, Fei and Huang, Yalou %D = 2009 %T = Social Tag Prediction Base on Supervised Ranking Model

@inproceedings{lipzcak2009tag, author = {Lipczak, Marek and Hu, Yeming and Kollet, Yael and Milios, Evangelos}, title = {Tag Sources for Recommendation in Collaborative Tagging Systems}, editor = {Eisterlehner, Folke and Hotho, Andreas and Jäschke, Robert}, booktitle = {ECML PKDD Discovery Challenge 2009 (DC09)}, series = {CEUR-WS.org}, year = {2009}, volume = {497}, pages = {157--172}, url = {http://ceur-ws.org/Vol-497/paper_19.pdf}, keywords = {challenge, dc09, recommender, tag} }

%0 = inproceedings %A = Lipczak, Marek and Hu, Yeming and Kollet, Yael and Milios, Evangelos %B = ECML PKDD Discovery Challenge 2009 (DC09) %D = 2009 %T = Tag Sources for Recommendation in Collaborative Tagging Systems %U = http://ceur-ws.org/Vol-497/paper_19.pdf

Tag recommendation is the task of predicting a personalized list of tags for a user given an item. This is important for many websites with tagging capabilities like last.fm or delicious. In this paper, we propose a method for tag recommendation based on tensor factorization (TF). In contrast to other TF methods like higher order singular value decomposition (HOSVD), our method RTF ('ranking with tensor factorization') directly optimizes the factorization model for the best personalized ranking. RTF handles missing values and learns from pairwise ranking constraints. Our optimization criterion for TF is motivated by a detailed analysis of the problem and of interpretation schemes for the observed data in tagging systems. In all, RTF directly optimizes for the actual problem using a correct interpretation of the data. We provide a gradient descent algorithm to solve our optimization problem. We also provide an improved learning and prediction method with runtime complexity analysis for RTF. The prediction runtime of RTF is independent of the number of observations and only depends on the factorization dimensions. Besides the theoretical analysis, we empirically show that our method outperforms other state-of-the-art tag recommendation methods like FolkRank, PageRank and HOSVD both in quality and prediction runtime.

@inproceedings{rendle2009learning, author = {Rendle, Steffen and Balby Marinho, Leandro and Nanopoulos, Alexandros and Schmidt-Thieme, Lars}, title = {Learning optimal ranking with tensor factorization for tag recommendation}, booktitle = {KDD '09: Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining}, publisher = {ACM}, address = {New York, NY, USA}, year = {2009}, pages = {727--736}, url = {http://portal.acm.org/citation.cfm?doid=1557019.1557100}, doi = {10.1145/1557019.1557100}, isbn = {978-1-60558-495-9}, keywords = {folksonomy, ranking, recommender, tag, tagging, tensor}, abstract = {Tag recommendation is the task of predicting a personalized list of tags for a user given an item. This is important for many websites with tagging capabilities like last.fm or delicious. In this paper, we propose a method for tag recommendation based on tensor factorization (TF). In contrast to other TF methods like higher order singular value decomposition (HOSVD), our method RTF ('ranking with tensor factorization') directly optimizes the factorization model for the best personalized ranking. RTF handles missing values and learns from pairwise ranking constraints. Our optimization criterion for TF is motivated by a detailed analysis of the problem and of interpretation schemes for the observed data in tagging systems. In all, RTF directly optimizes for the actual problem using a correct interpretation of the data. We provide a gradient descent algorithm to solve our optimization problem. We also provide an improved learning and prediction method with runtime complexity analysis for RTF. The prediction runtime of RTF is independent of the number of observations and only depends on the factorization dimensions. Besides the theoretical analysis, we empirically show that our method outperforms other state-of-the-art tag recommendation methods like FolkRank, PageRank and HOSVD both in quality and prediction runtime.} }

%0 = inproceedings %A = Rendle, Steffen and Balby Marinho, Leandro and Nanopoulos, Alexandros and Schmidt-Thieme, Lars %B = KDD '09: Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining %C = New York, NY, USA %D = 2009 %I = ACM %T = Learning optimal ranking with tensor factorization for tag recommendation %U = http://portal.acm.org/citation.cfm?doid=1557019.1557100

@inproceedings{xi2009content, author = {Si, Xiance and Liu, Zhiyuan and Li, Peng and Jiang, Qixia and Sun, Maosong}, title = {Content-based and Graph-based Tag Suggestion}, editor = {Eisterlehner, Folke and Hotho, Andreas and Jäschke, Robert}, series = {CEUR-WS.org}, year = {2009}, volume = {497}, pages = {243--260}, keywords = {dc09, recommender, tag} }

%0 = inproceedings %A = Si, Xiance and Liu, Zhiyuan and Li, Peng and Jiang, Qixia and Sun, Maosong %D = 2009 %T = Content-based and Graph-based Tag Suggestion

Social bookmarking systems allow users to organise collections of resources on the Web in a collaborative fashion. The increasing popularity of these systems as well as first insights into their emergent semantics have made them relevant to disciplines like knowledge extraction and ontology learning. The problem of devising methods to measure the semantic relatedness between tags and characterizing it semantically is still largely open. Here we analyze three measures of tag relatedness: tag co-occurrence, cosine similarity of co-occurrence distributions, and FolkRank, an adaptation of the PageRank algorithm to folksonomies. Each measure is computed on tags from a large-scale dataset crawled from the social bookmarking system del.icio.us. To provide a semantic grounding of our findings, a connection to WordNet (a semantic lexicon for the English language) is established by mapping tags into synonym sets of WordNet, and applying there well-known metrics of semantic similarity. Our results clearly expose different characteristics of the selected measures of relatedness, making them applicable to different subtasks of knowledge extraction such as synonym detection or discovery of concept hierarchies.

@inproceedings{cattuto2008semantic, author = {Cattuto, Ciro and Benz, Dominik and Hotho, Andreas and Stumme, Gerd}, title = {Semantic Analysis of Tag Similarity Measures in Collaborative Tagging Systems}, booktitle = {Proceedings of the 3rd Workshop on Ontology Learning and Population (OLP3)}, address = {Patras, Greece}, year = {2008}, pages = {39--43}, url = {http://olp.dfki.de/olp3/}, isbn = {978-960-89282-6-8}, keywords = {2008, folksonomy, learning, ol_tut2010, ontology, similarity, tag, webzu}, abstract = {Social bookmarking systems allow users to organise collections of resources on the Web in a collaborative fashion. The increasing popularity of these systems as well as first insights into their emergent semantics have made them relevant to disciplines like knowledge extraction and ontology learning. The problem of devising methods to measure the semantic relatedness between tags and characterizing it semantically is still largely open. Here we analyze three measures of tag relatedness: tag co-occurrence, cosine similarity of co-occurrence distributions, and FolkRank, an adaptation of the PageRank algorithm to folksonomies. Each measure is computed on tags from a large-scale dataset crawled from the social bookmarking system del.icio.us. To provide a semantic grounding of our findings, a connection to WordNet (a semantic lexicon for the English language) is established by mapping tags into synonym sets of WordNet, and applying there well-known metrics of semantic similarity. Our results clearly expose different characteristics of the selected measures of relatedness, making them applicable to different subtasks of knowledge extraction such as synonym detection or discovery of concept hierarchies.} }

%0 = inproceedings %A = Cattuto, Ciro and Benz, Dominik and Hotho, Andreas and Stumme, Gerd %B = Proceedings of the 3rd Workshop on Ontology Learning and Population (OLP3) %C = Patras, Greece %D = 2008 %T = Semantic Analysis of Tag Similarity Measures in Collaborative Tagging Systems %U = http://olp.dfki.de/olp3/

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.

@article{jaeschke2008tag, author = {Jäschke, Robert and Marinho, Leandro and Hotho, Andreas and Schmidt-Thieme, Lars and Stumme, Gerd}, title = {Tag Recommendations in Social Bookmarking Systems}, editor = {Giunchiglia, Enrico}, journal = {AI Communications}, publisher = {IOS Press}, address = {Amsterdam}, year = {2008}, volume = {21}, number = {4}, pages = {231--247}, url = {http://www.kde.cs.uni-kassel.de/pub/pdf/jaeschke2008tag.pdf}, doi = {10.3233/AIC-2008-0438}, issn = {0921-7126}, keywords = {2008, myown, recommender, tag, top, webzu}, 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. } }

%0 = article %A = Jäschke, Robert and Marinho, Leandro and Hotho, Andreas and Schmidt-Thieme, Lars and Stumme, Gerd %C = Amsterdam %D = 2008 %I = IOS Press %T = Tag Recommendations in Social Bookmarking Systems %U = http://www.kde.cs.uni-kassel.de/pub/pdf/jaeschke2008tag.pdf

This paper aims to quantify two common assumptions about social tagging: (1) that tags are "meaningful" and (2) that the tagging process is influenced by tag suggestions. For (1), we analyze the semantic properties of tags and the relationship between the tags and the content of the tagged page. Our analysis is based on a corpus of search keywords, contents, titles, and tags applied to several thousand popular Web pages. Among other results, we find that the more popular tags of a page tend to be the more meaningful ones. For (2), we develop a model of how the influence of tag suggestions can be measured. From a user study with over 4,000 participants, we conclude that roughly one third of the tag applications may be induced by the suggestions. Our results would be of interest for designers of social tagging systems and are a step towards understanding how to best leverage social tags for applications such as search and information extraction.

@inproceedings{suchanek2008social, author = {Suchanek, Fabian M. and Vojnovic, Milan and Gunawardena, Dinan}, title = {Social tags: meaning and suggestions}, booktitle = {CIKM '08: Proceeding of the 17th ACM conference on Information and knowledge management}, publisher = {ACM}, address = {New York, NY, USA}, year = {2008}, pages = {223--232}, url = {http://portal.acm.org/citation.cfm?id=1458114}, doi = {http://doi.acm.org/10.1145/1458082.1458114}, isbn = {978-1-59593-991-3}, keywords = {folksonomy, meaning, recommender, tag}, abstract = {This paper aims to quantify two common assumptions about social tagging: (1) that tags are "meaningful" and (2) that the tagging process is influenced by tag suggestions. For (1), we analyze the semantic properties of tags and the relationship between the tags and the content of the tagged page. Our analysis is based on a corpus of search keywords, contents, titles, and tags applied to several thousand popular Web pages. Among other results, we find that the more popular tags of a page tend to be the more meaningful ones. For (2), we develop a model of how the influence of tag suggestions can be measured. From a user study with over 4,000 participants, we conclude that roughly one third of the tag applications may be induced by the suggestions. Our results would be of interest for designers of social tagging systems and are a step towards understanding how to best leverage social tags for applications such as search and information extraction.} }

%0 = inproceedings %A = Suchanek, Fabian M. and Vojnovic, Milan and Gunawardena, Dinan %B = CIKM '08: Proceeding of the 17th ACM conference on Information and knowledge management %C = New York, NY, USA %D = 2008 %I = ACM %T = Social tags: meaning and suggestions %U = http://portal.acm.org/citation.cfm?id=1458114

ConTag is an approach to generate semantic tag recommendations for documents
sed on Semantic Web ontologies and Web 2.0 services. We designed and implemented a process to normalize documents to RDF format, extract document topics using Web 2.0 services and finally match extracted topics to a Semantic
b ontology.

e to ConTag we are able to show that the information provided by Web 2.0 services in combination with a Semantic Web ontology enables the generation of relevant semantic tag recommendations for documents. The main contribution of this work is a semantic tag recommendation process based on a choreography of Web
0 services.

@inproceedings{adrian2007contag, author = {Adrian, Benjamin and Sauermann, Leo and Roth-Berghofer, Thomas}, title = {ConTag: A semantic tag recommendation system}, editor = {Pellegrini, Tassilo and Schaffert, Sebastian}, booktitle = {Proceedings of I-Semantics' 07}, publisher = {JUCS}, year = {2007}, pages = {297-304}, url = {http://www.dfki.uni-kl.de/~sauermann/papers/horak+2007a.pdf}, keywords = {contag, recommender, semantic, tag, web}, abstract = {ConTag is an approach to generate semantic tag recommendations for documents
sed on Semantic Web ontologies and Web 2.0 services. We designed and implemented a process to normalize documents to RDF format, extract document topics using Web 2.0 services and finally match extracted topics to a Semantic
b ontology.

e to ConTag we are able to show that the information provided by Web 2.0 services in combination with a Semantic Web ontology enables the generation of relevant semantic tag recommendations for documents. The main contribution of this work is a semantic tag recommendation process based on a choreography of Web
0 services.} }

%0 = inproceedings %A = Adrian, Benjamin and Sauermann, Leo and Roth-Berghofer, Thomas %B = Proceedings of I-Semantics' 07 %D = 2007 %I = JUCS %T = ConTag: A semantic tag recommendation system %U = http://www.dfki.uni-kl.de/~sauermann/papers/horak+2007a.pdf

Tag clouds provide an aggregate of tag-usage statistics. They are typically sent as in-line HTML to browsers. However, display mechanisms suited for ordinary text are not ideal for tags, because font sizes may vary widely on a line. As well, the typical layout does not account for relationships that may be known between tags. This paper presents models and algorithms to improve the display of tag clouds that con- sist of in-line HTML, as well as algorithms that use nested tables to achieve a more general 2-dimensional layout in which tag relationships are considered. The first algorithms leverage prior work in typesetting and rectangle packing, whereas the second group of algorithms leverage prior work in Electronic Design Automation. Experiments show our algorithms can be efficiently implemented and perform well.

@article{kaser2007tagcloud, author = {Kaser, Owen and Lemire, Daniel}, title = {Tag-Cloud Drawing: Algorithms for Cloud Visualization}, journal = {CoRR}, year = {2007}, volume = {abs/cs/0703109}, note = {informal publication}, url = {http://arxiv.org/abs/cs/0703109}, keywords = {cloud, drawing, folksonomy, tag, visualization}, abstract = {Tag clouds provide an aggregate of tag-usage statistics. They are typically sent as in-line HTML to browsers. However, display mechanisms suited for ordinary text are not ideal for tags, because font sizes may vary widely on a line. As well, the typical layout does not account for relationships that may be known between tags. This paper presents models and algorithms to improve the display of tag clouds that con- sist of in-line HTML, as well as algorithms that use nested tables to achieve a more general 2-dimensional layout in which tag relationships are considered. The first algorithms leverage prior work in typesetting and rectangle packing, whereas the second group of algorithms leverage prior work in Electronic Design Automation. Experiments show our algorithms can be efficiently implemented and perform well. } }

%0 = article %A = Kaser, Owen and Lemire, Daniel %D = 2007 %T = Tag-Cloud Drawing: Algorithms for Cloud Visualization %U = http://arxiv.org/abs/cs/0703109

@misc{illig2006entwurf, author = {Illig, Jens}, title = {Entwurf und Integration eines Item-Based Collaborative Filtering
g Recommender Systems in das BibSonomy-Projekt}, year = {2006}, url = {http://www.kde.cs.uni-kassel.de/lehre/arbeiten/documents/illig2006entwurf.pdf}, keywords = {2006, bibsonomy, kde-abschlussarbeit, projekt, recommender, tag} }

%0 = misc %A = Illig, Jens %D = 2006 %T = Entwurf und Integration eines Item-Based Collaborative Filtering
g Recommender Systems in das BibSonomy-Projekt %U = http://www.kde.cs.uni-kassel.de/lehre/arbeiten/documents/illig2006entwurf.pdf

Folksonomies are classification schemes that emerge from the collective actions of users who tag resources with an unrestricted
t of key terms. There has been a flurry of activity in this domain recently with a number of high profile web sites andsearch engines adopting the practice. They have sparked a great deal of excitement and debate in the popular and technicalliterature, accompanied by a number of analyses of the statistical properties of tagging behavior. However, none has addressedthe deep nature of folksonomies. What is the nature of a tag? Where does it come from? How is it related to a resource? Inthis paper we present a study in which the linguistic properties of folksonomies reveal them to contain, on the one hand,tags that are similar to standard categories in taxonomies. But on the other hand, they contain additional tags to describeclass properties. The implications of the findings for the relationship between folksonomy and ontology are discussed.

@inproceedings{veres2006language, author = {Veres, Csaba}, title = {The Language of Folksonomies: What Tags Reveal About User Classification}, editor = {Kop, Christian and Fliedl, Günther and Mayr, Heinrich C. and Métais, Elisabeth}, booktitle = {Natural Language Processing and Information Systems}, series = {Lecture Notes in Computer Science}, publisher = {Springer}, address = {Berlin/Heidelberg}, year = {2006}, volume = {3999}, pages = {58--69}, url = {http://dx.doi.org/10.1007/11765448_6}, doi = {10.1007/11765448}, isbn = {978-3-540-34616-6}, keywords = {classification, folksonomy, language, tag, tagging, user}, abstract = {Folksonomies are classification schemes that emerge from the collective actions of users who tag resources with an unrestricted
t of key terms. There has been a flurry of activity in this domain recently with a number of high profile web sites andsearch engines adopting the practice. They have sparked a great deal of excitement and debate in the popular and technicalliterature, accompanied by a number of analyses of the statistical properties of tagging behavior. However, none has addressedthe deep nature of folksonomies. What is the nature of a tag? Where does it come from? How is it related to a resource? Inthis paper we present a study in which the linguistic properties of folksonomies reveal them to contain, on the one hand,tags that are similar to standard categories in taxonomies. But on the other hand, they contain additional tags to describeclass properties. The implications of the findings for the relationship between folksonomy and ontology are discussed.} }

%0 = inproceedings %A = Veres, Csaba %B = Natural Language Processing and Information Systems %C = Berlin/Heidelberg %D = 2006 %I = Springer %T = The Language of Folksonomies: What Tags Reveal About User Classification %U = http://dx.doi.org/10.1007/11765448_6

Content organization over the Internet went through several
teresting phases of evolution: from structured directories to
structured Web search engines and more recently, to tagging
a way for aggregating information, a step towards the
mantic web vision. Tagging allows ranking and data
ganization to directly utilize inputs from end users, enabling
chine processing of Web content. Since tags are created by
dividual users in a free form, one important problem facing
gging is to identify most appropriate tags, while eliminating
ise and spam. For this purpose, we define a set of general
iteria for a good tagging system. These criteria include high
verage of multiple facets to ensure good recall, least effort to
duce the cost involved in browsing, and high popularity to
sure tag quality. We propose a collaborative tag suggestion
gorithm using these criteria to spot high-quality tags. The
oposed algorithm employs a goodness measure for tags derived
om collective user authorities to combat spam. The goodness
asure is iteratively adjusted by a reward-penalty algorithm,
ich also incorporates other sources of tags, e.g., content-based
to-generated tags. Our experiments based on My Web 2.0 show
at the algorithm is effective.

@inproceedings{xu2006tsw, author = {Xu, Z. and Fu, Y. and Mao, J. and Su, D.}, title = {Towards the semantic web: Collaborative tag suggestions}, booktitle = {Proceedings of the Collaborative Web Tagging Workshop at the WWW 2006}, address = {Edinburgh, Scotland}, year = {2006}, url = {http://www.ibiblio.org/www_tagging/2006/13.pdf}, keywords = {recommender, tag, tagging}, abstract = {Content organization over the Internet went through several
teresting phases of evolution: from structured directories to
structured Web search engines and more recently, to tagging
a way for aggregating information, a step towards the
mantic web vision. Tagging allows ranking and data
ganization to directly utilize inputs from end users, enabling
chine processing of Web content. Since tags are created by
dividual users in a free form, one important problem facing
gging is to identify most appropriate tags, while eliminating
ise and spam. For this purpose, we define a set of general
iteria for a good tagging system. These criteria include high
verage of multiple facets to ensure good recall, least effort to
duce the cost involved in browsing, and high popularity to
sure tag quality. We propose a collaborative tag suggestion
gorithm using these criteria to spot high-quality tags. The
oposed algorithm employs a goodness measure for tags derived
om collective user authorities to combat spam. The goodness
asure is iteratively adjusted by a reward-penalty algorithm,
ich also incorporates other sources of tags, e.g., content-based
to-generated tags. Our experiments based on My Web 2.0 show
at the algorithm is effective.} }

%0 = inproceedings %A = Xu, Z. and Fu, Y. and Mao, J. and Su, D. %B = Proceedings of the Collaborative Web Tagging Workshop at the WWW 2006 %C = Edinburgh, Scotland %D = 2006 %T = Towards the semantic web: Collaborative tag suggestions %U = http://www.ibiblio.org/www_tagging/2006/13.pdf

Recommender systems apply knowledge discovery techniques to the problem of making personalized recommendations for information, products or services during a live interaction. These systems, especially the k-nearest neighbor collaborative filtering based ones, are achieving widespread success on the Web. The tremendous growth in the amount of available information and the number of visitors to Web sites in recent years poses some key challenges for recommender systems. These are: producing high quality recommendations, performing many recommendations per second for millions of users and items and achieving high coverage in the face of data sparsity. In traditional collaborative filtering systems the amount of work increases with the number of participants in the system. New recommender system technologies are needed that can quickly produce high quality recommendations, even for very large-scale problems. To address these issues we have explored item-based collaborative filtering techniques. Item-based techniques first analyze the user-item matrix to identify relationships between different items, and then use these relationships to indirectly compute recommendations for users. In this paper we analyze different item-based recommendation generation algorithms. We look into different techniques for computing item-item similarities (e.g., item-item correlation vs. cosine similarities between item vectors) and different techniques for obtaining recommendations from them (e.g., weighted sum vs. regression model). Finally, we experimentally evaluate our results and compare them to the basic k-nearest neighbor approach. Our experiments suggest that item-based algorithms provide dramatically better performance than user-based algorithms, while at the same time providing better quality than the best available user-based algorithms.

@inproceedings{sarwar2001item, author = {Sarwar, Badrul and Karypis, George and Konstan, Joseph and Riedl, John}, title = {Item-based collaborative filtering recommendation algorithms}, booktitle = {WWW '01: Proceedings of the 10th International Conference on World Wide Web}, publisher = {ACM}, address = {New York, NY, USA}, year = {2001}, pages = {285--295}, url = {http://portal.acm.org/citation.cfm?id=372071}, doi = {10.1145/371920.372071}, isbn = {1-58113-348-0}, keywords = {collaborative, filtering, recommender, stair, tag}, abstract = {Recommender systems apply knowledge discovery techniques to the problem of making personalized recommendations for information, products or services during a live interaction. These systems, especially the k-nearest neighbor collaborative filtering based ones, are achieving widespread success on the Web. The tremendous growth in the amount of available information and the number of visitors to Web sites in recent years poses some key challenges for recommender systems. These are: producing high quality recommendations, performing many recommendations per second for millions of users and items and achieving high coverage in the face of data sparsity. In traditional collaborative filtering systems the amount of work increases with the number of participants in the system. New recommender system technologies are needed that can quickly produce high quality recommendations, even for very large-scale problems. To address these issues we have explored item-based collaborative filtering techniques. Item-based techniques first analyze the user-item matrix to identify relationships between different items, and then use these relationships to indirectly compute recommendations for users. In this paper we analyze different item-based recommendation generation algorithms. We look into different techniques for computing item-item similarities (e.g., item-item correlation vs. cosine similarities between item vectors) and different techniques for obtaining recommendations from them (e.g., weighted sum vs. regression model). Finally, we experimentally evaluate our results and compare them to the basic k-nearest neighbor approach. Our experiments suggest that item-based algorithms provide dramatically better performance than user-based algorithms, while at the same time providing better quality than the best available user-based algorithms.} }

%0 = inproceedings %A = Sarwar, Badrul and Karypis, George and Konstan, Joseph and Riedl, John %B = WWW '01: Proceedings of the 10th International Conference on World Wide Web %C = New York, NY, USA %D = 2001 %I = ACM %T = Item-based collaborative filtering recommendation algorithms %U = http://portal.acm.org/citation.cfm?id=372071