With the increased popularity of Web 2.0 services in the last years data privacy has become a major concern for users. The more personal data users reveal, the more difficult it becomes to control its disclosure in the web. However, for Web 2.0 service providers, the data provided by users is a valuable source for offering effective, personalised data mining services. One major application is the detection of spam in social bookmarking systems: in order to prevent a decrease of content quality, providers need to distinguish spammers and exclude them from the system. They thereby experience a conflict of interests: on the one hand, they need to identify spammers based on the information they collect about users, on the other hand, they need to respect privacy concerns and process as few personal data as possible. It would therefore be of tremendous help for system developers and users to know which personal data are needed for spam detection and which can be ignored. In this paper we address these questions by presenting a data privacy aware feature engineering approach. It consists of the design of features for spam classification which are evaluated according to both, performance and privacy conditions. Experiments using data from the social bookmarking system BibSonomy show that both conditions must not exclude each other.

The annotation of web sites in social bookmarking systemshas become a popular way to manage and find informationon the web. The community structure of such systems attractsspammers: recent post pages, popular pages or specifictag pages can be manipulated easily. As a result, searchingor tracking recent posts does not deliver quality resultsannotated in the community, but rather unsolicited, oftencommercial, web sites. To retain the benefits of sharingone’s web content, spam-fighting mechanisms that can facethe flexible strategies of spammers need to be developed.

As the number of resources on the web exceeds by far the number of
cuments one can track, it becomes increasingly difficult to remain
to date on ones own areas of interest. The problem becomes more
vere with the increasing fraction of multimedia data, from which
is difficult to extract some conceptual description of their
ntents.

ne way to overcome this problem are social bookmark tools, which
e rapidly emerging on the web. In such systems, users are setting
lightweight conceptual structures called folksonomies, and
ercome thus the knowledge acquisition bottleneck. As more and more
ople participate in the effort, the use of a common vocabulary
comes more and more stable. We present an approach for discovering
pic-specific trends within folksonomies. It is based on a
fferential adaptation of the PageRank algorithm to the triadic
pergraph structure of a folksonomy. The approach allows for any
nd of data, as it does not rely on the internal structure of the
cuments. In particular, this allows to consider different data
pes in the same analysis step. We run experiments on a large-scale
al-world snapshot of a social bookmarking system.

Ein wichtiger Baustein des neu entdeckten World Wide Web -- des "`Web 2.0"' -- stellen
lksonomies dar. In diesen Systemen können Benutzer gemeinsam Ressourcen verwalten und
t Schlagwörtern versehen. Die dadurch entstehenden begrifflichen Strukturen stellen
n interessantes Forschungsfeld dar. Dieser Artikel untersucht Ansätze und Wege zur
tdeckung und Strukturierung von Nutzergruppen ("Communities") in Folksonomies.