Querying Multimedia Documents By Spatiotemporal Structure

Interactive multimedia documents are rapidly becoming available on the WWW. Navigating in such large document repositories requires flexible retrieval techniques based on spatiotemporal structure. In this paper we address this issue by extending the notion of conceptual neighborhood (a concept describing similarity among relations between time intervals) to various resolution levels and higher dimensions. We propose a binary string encoding of relations which allows the automatic derivation of similarity measures and apply our framework for effective support of multimedia queries.