Heterogeneous Data Co-Clustering by Pseudo-Semantic Affinity Functions.
Alberto Messina, Maurizio Montagnuolo · 2011
Abstract. The convergence between Web technology and multimedia production is enabling the distribution of content through dynamic media platforms such as RSS feeds and hybrid digital television. Heterogeneous data clustering is needed to analyse, manage and access desired information from this variety of information sources. This paper defines a new class of pseudo-semantic affinity functions that allow for a compact representation of cross-modal documents relations. Keywords: Co-clustering, Pseudo-semantic Model, Syntactic Affinity 1 Introduction and Related Work The proliferation of multimedia production tools is enabling the convergence of different media technologies. The result of this technological breakthrough is the creation of new relationships between different media objects and the modalities in which they are generated and consumed.