The HERMES Search Engine

The rapid and wide diffusion of multimedia information published on the World Wide Web makes particularly relevant the problem of efficient and effective content-based retrieval of multimedia Web pages. Most of the existing search engines base the retrieval only on text while few research prototype systems support a retrieval based on the physical properties of multimedia objects.

There is an emerging need of a new generation of search engines that try to exploit the full multimedia information present on the Web.

We propose an approach which makes possible to retrieve Web pages by using their entire content: text, physical properties of multimedia objects as well as semantic information they represent, information about the "context" in which a multimedia object appears. The proposed approach exploits a data model for content based retrieval of multimedia data to describe the various multimedia components, their structure and their relationships with a pre-defined taxonomy of concepts.

If you are interested in more detail on the HERMES Search Engine, you can find some papers that describe the used approach:

• "Supporting image search on the web", G.Amato, F. Rabitti, P. Savino, The challenge of Image Retrieval-Workshop on Image Retrieval , February 5, 1998, University of Northumbria at Newcastle, UK.
  Paper (gzipped postscript 497 Kb).

• "Multimedia Document Search on the Web" G.Amato, F.Rabitti, P. Savino, Seventh International World Wide Web Conference (WWW7), Brisbane, Queensland, Australia 14 - 18 April 1998.
  Paper (HTML)
  Slides (HTML)