Alexander Höhfeld

Self-organizing collaborative filtering in global-scale massive multi-user virtual environments

Due to the huge amount of available information in todays society, it becomes more and more difficult for the consumer to locate the most useful information for a specific topic. Recommender systems using collaborative filtering (CF) are a popular technique for reducing information overload and finding useful information on the Internet. However, in massive global-scale multi-user virtual environments different approaches are required from those used within the currently dominant centralized infrastructures or lately investigated P2P approaches. Within this paper we present a novel collaborative filtering algorithm used within the HyperVerse -- a P2P-based self-organizing middleware service for massively distributed virtual worlds -- to generate and manage recommendations for HyperVerse object favorites. Due to its global extent considering users and possible ratings, using a monolithic database-backed recommendation service or huge profile- or item-rating-matrices does not scale in our scenario. The decentralized approach presented within this paper creates per user ratings in an adaptive and transparent way by comparing public favorites of passer-by users with personal peer data, weighted by self-adjusting buddy lists.

SimCon - a simulation and visualization environment for overlay networks and large-scale applications

A substantial number of massive large-scale applications require scalable underlying network topologies. Nowadays structured Peer-to-Peer overlay networks meet these requirements very well. But there is still a need to decide which of these overlay networks is most suitable for providing the best possible performance for a certain application. This paper describes SimCon - a simulation environment for overlay networks and large-scale applications. SimCon allows the comparison of different overlay networks with respect to predefined metrics derived from requirements of the considered application. This approach allows determining which overlay network meets the needs of a given application best, which in turn is a great support for developers of large-scale applications.

HiOPS overlay - efficient provision of multicast in peer-to-peer systems

During the last few years the application scope of peer-to-peer systems has been notably extended. While the problems of scalable data localization have been exhaustively addressed, one may notice that the problem of multicast cost and latency delays in very large, global scale peer-to-peer networks still remains inadequately considered. Many large scale applications will suffer from this multicast cost if the number of users increases. In order to address this problem, we introduce HiOPS - a hierarchical two tier peer-to-peer system which provides a reasonable tradeoff between minimization of multicast cost and latency delays and reduction of maintenance and management cost. Because of its properties and characteristics HiOPS can be considered as a basic infrastructure for a wide range of peer-to-peer applications.