Knowledge-based Indexing and Retrieval of Multimedia InformaTion

Short description

The research question addressed in this project is: what methods, techniques and tools can be developed for the description and indexing of multimedia information, such that efficient and effective retrieval of information can be achieved in a distributed, multilingual and multipurpose environment. The approach taken in this project is that of knowledge-based indexing and retrieval: an element of information -e.g. a small video fragment- is described using a number of descriptive dimensions (descriptors) and values taken from a set of standardised thesauri (ontologies). Each ontology takes the form of a structured knowledge base that covers certain aspects of the information domain. The explicit representation of the viewpoints on which the ontology is based, allows the development of  tools that support non-expert users, with varying goals, in query formulation and search. In addition, ontology technology provides the handles to access multiple databases with different indexing schemes, through mapping operations.

People involved

SWI, University of Amsterdam

prof. dr. B.J. Wielinga  [
dr. B. Bredeweg [

IMSE, Vrije Universiteit, Amsterdam

prof. dr. J.M. Akkermans  []
dr. D. Fensel [
drs. M.C.A. Klein [