Knowledge modeling for enhanced information retrieval and visualization

Abstract

The advent of technologies in information retrieval driven by users' requests calls for an effort to annotate multimedia contents. The exponential growth of digital resources makes the use of manual solutions unfeasible. In recent years domain-specific ontologies (abstract model of some domain) have been adopted to enhance content based annotation and retrieval of these digital resources. In this paper we present an ontology that models implicit and explicit knowledge and information conveyed within a conference life cycle. Our model, presented as a High-level modEL for cOnference (HELO), is used on the one hand to perform high level annotation of video recordings of a conference allowing granular search facilities and complex queries, and on the other hand to enhance knowledge retrieval and visualization. As a proof-of-concept HELO has been integrated into CALIMERA (Conference Advanced Level Information ManagEment & RetrievAl.) a framework we developed that handles conference information. © Springer-Verlag Berlin Heidelberg 2010.