Over the last years the means to collect personal data implicitly or explicitly over the Web and by various kinds of sensors and to combine data for profiling individuals has dram...
Abstract—Parallel file systems are designed to mask the everincreasing gap between CPU and disk speeds via parallel I/O processing. While they have become an indispensable compo...
We present a content-driven reputation system for Wikipedia authors. In our system, authors gain reputation when the edits they perform to Wikipedia articles are preserved by subs...
The collaborative filtering approach to recommender systems predicts user preferences for products or services by learning past useritem relationships. In this work, we propose no...
In this paper, we design recommender systems for weblogs based on the link structure among them. We propose algorithms based on refined random walks and spectral methods. First, w...