In this paper, we look at the "social tag prediction" problem. Given a set of objects, and a set of tags applied to those objects by users, can we predict whether a given tag could/should be applied to a particular object? We investigated this question using one of the largest crawls of the social bookmarking system del.icio.us gathered to date. For URLs in del.icio.us, we predicted tags based on page text, anchor text, surrounding hosts, and other tags applied to the URL. We found an entropy-based metric which captures the generality of a particular tag and informs an analysis of how well that tag can be predicted. We also found that tag-based association rules can produce very high-precision predictions as well as giving deeper understanding into the relationships between tags. Our results have implications for both the study of tagging systems as potential information retrieval tools, and for the design of such systems.

While recommender systems tell users what items they might like, explanations of recommendations reveal why they might like them. Explanations provide many benefits, from improving user satisfaction to helping users make better decisions. This paper introduces tagsplanations, which are explanations based on community tags. Tagsplanations have two key components: tag relevance, the degree to which a tag describes an item, and tag preference, the user's sentiment toward a tag. We develop novel algorithms for estimating tag relevance and tag preference, and we conduct a user study exploring the roles of tag relevance and tag preference in promoting effective tagsplanations. We also examine which types of tags are most useful for tagsplanations.

Empfehlungssysteme tragen Inhalte individuell an Nutzer im WWW heran, basierend auf deren konkreten Bedürfnissen, Vorlieben und Interessen. Solche Systeme können Produkte, Services, Nutzer (mit analogen Interessen) uvm. vorschlagen und stellen damit – gerade im Web 2.0-Zeitalter – eine besondere Form der Personalisierung sowie des social networking dar. Damit bieten Empfehlungssysteme Anbietern im ECommerce einen entscheidenden Marktvorteil, weshalb die Auswertung der Kundendaten bei großen Firmen wie Amazon, Google oder Ebay eine hohe Priorität besitzt. Aus diesem Grund wird im vorliegenden Artikel auf die Ansätze von Empfehlungssystemen, welche auf unterschiedliche Weise die Bedürfnisse des Nutzers aufgreifen bzw. „vorausahnen“ und ihm Vorschläge (aus verschiedenen Bereichen) unterbreiten können, eingegangen. Der Artikel liefert eine Definition und Darstellung der Arbeitsweisen von Empfehlungssystemen. Dabei werden die verschiedenen Methodiken jener Dienste vergleichend erläutert, um ihre jeweiligen Vor- und Nachteile deutlich zu machen. Außerdem wird der Ontologie- und Folksonomy-Einsatz innerhalb von Empfehlungssystemen beleuchtet, um Chancen und Risiken der Anwendung von Methoden der Wissensrepräsentation für zukünftige Forschungsarbeiten einschätzen zu können. Recommender Systems in an Information Science View – The State of the Art Recommender systems offer content individually to users in the WWW, based on their concrete needs, preferences and interests. Those systems can propose products, services, users (with analogous interests), etc.) and represent a special form of personalisation as well as of social networking – exactly in the Web 2.0 age. Recommender systems offer e.g. suppliers in the e-commerce a crucial market advantage. So, the evaluation of the customer data has high priority at big companies like Amazon, Google or Ebay. For this reason we engaged in recommender systems, which take up the user’s needs in different ways, to “anticipate“ needs and make suggestions (from different areas) to the user. This review article achieves a definition and representation of operations and methods of recommender systems. Exactly the different methodologies of those services should be expounded comparativly on that occasion in order to represent advantages and disadvantages. The use of ontologies and folksonomies as implementations in recommender systems is portrayed in order to be able to take into consideration chances and risks of the application of knowledge representation methods for future researches.

Personalized recommendation is used to conquer the information overload problem, and collaborative filtering recommendation (CF) is one of the most successful recommendation techniques to date. However, CF becomes less effective when users have multiple interests, because users have similar taste in one aspect may behave quite different in other aspects. Information got from social bookmarking websites not only tells what a user likes, but also why he or she likes it. This paper proposes a division algorithm and a CubeSVD algorithm to analysis this information, distill the interrelations between different users’ various interests, and make better personalized recommendation based on them. Experiment reveals the superiority of our method over traditional CF methods. ER -