Recent growth of social classification systems due to steadily increasing popularity has established a multitude of heterogeneous isolated, non-integrated, and non-interoperable tag spaces. Contrary to current research predominantly focusing on single folksonomies, we exploit cross-space similarities to improve a variety of tagging use cases beyond the limits of one folksonomy. This paper presents the results of practical studies concerning cross-space analysis of (co-)tag spaces of five well-established social classification services for tagging of bookmarks (del.icio.us, BibSonomy bookmarks), and publications (BibSonomy publications, CiteULike, Connotea). The studies are based on one month data sets of RSS recent feeds from the same time scope. We provide a profound motivation for cross-space tagging, and give insight into similarities and intersections of (top ranking) (co-)tag spaces as well as convergence aspects over time. Categories and Subject Descriptors H.3 [Information Stor...
Steffen Oldenburg, Martin Garbe, Clemens H. Cap