Exploring semantic relationships across internet resources

As the social character of the Web grows stronger, online multimedia resources become more tangled, exhibiting multiple and complex relationships. Often, these relationships are not easily spotted by search engines because they are not explicit. Still, when searching for content, users would certainly appreciate to receive results indicating all resources that are somehow related to their topic of interest without having to explicitly request for it. In spite of this trend, current technology for declaring and describing resources on the Internet as well as their inter-relationships is very much fragmented and the most commonly used standards, such as HTML, do not offer adequate functionality. While the goal of the MPEG-21 standard is to provide an all encompassing, universal and standard way to perform the declaration of digital resources, the work being developed in the IST CONVERGENCE project has shown that this standard, (specifically its part 3, Digital Item Identification), also has a limited ability to specify relationships between digital resources. This document presents a proposed extension to MPEG-21 Digital Item Identification, which enables it to support a uniform, semantically expressive and precise specification of the relationships between MPEG-21 digital items, rendering it a much more effective tool for a ubiquitous representation and interconnection of digital resources.