Integrating structured metadata with relational affinity propagation

Structured and semi-structured data describing entities, tax- onomies and ontologies appears in many domains. There is a huge interest in integrating structured information from multiple sources; however integrating structured data to in- fer complex common structures is a difficult task because the integration must aggregate similar structures while avoiding structural inconsistencies that may appear when the data is combined. In this work, we study the integration of struc- tured social metadata: shallow personal hierarchies specified by many individual users on the Social Web, and focus on in- ferring a collection of integrated, consistent taxonomies. We frame this task as an optimization problem with structural constraints. We propose a new inference algorithm, which we refer to as Relational Affinity Propagation (RAP) that ex- tends affinity propagation (Frey and Dueck 2007) by intro- ducing structural constraints. We validate the approach on a real-world social media dataset, collected from the photoshar- ing website Flickr. Our empirical results show that our pro- posed approach is able to construct deeper and denser struc- tures compared to an approach using only the standard affin- ity propagation algorithm.