MetaOn - Ontology Driven Metadata Construction and Management for Intelligent Search in Text and Image Collections

Abstract

State of the art in multimedia technology focuses in managing data collected from various sources, including documents, images, video, and speech. Therefore the effective management, analysis and mining of such heterogeneous data require the combination of various techniques. In this paper, we present an overview of the funded MetaOn project. The core objective of MetaOn is to construct and integrate semantically rich metadata collections extracted from documents, images and linguistic resources, to facilitate intelligent search and analysis. The proposed MetaOn framework involves ontology-based information extraction and data mining, semi-automatic construction of domain specific ontologies, content-based image indexing and retrieval, and metadata management. The Hellenic history has been chosen as a challenging application case study