NeuroCOLT

Neural Networks and Computational Learning Theory

 

About NeuroCOLT

Papers Archive

Research Areas

Partners

Coordinator

Events

info@neurocolt.org

 

NeuroCOLT workshop
on
Applications of Learning to Text and Images
Windsor, 30 April - 2 May 2001
Cumberland Lodge

A multiSOM approach exploiting synergy between text and image for information discovery in a multimedia context

Jean-Charles Lamirel

lamirel@loria.fr

A lot of experiments have shown that images, graphics and iconographic resources, thanks to their explanatory power, can be considered as a very fundamental component of a man-machine interface. The goal of our approach is to make use of this oustanding property of the images in order to provide a Digital Library with Information Discovering capabilities. The main tool that will be presented in this paper is the MicroNOMAD Discovering Tool. Its most important characteristic is both to provide user with emergent and Ğeasy to useğ analyses of an iconographic database content and with overall querying and browsing guidelines through the use of an advanced topographic interface model. Conversely to a lot of other more classical models, the MicroNOMAD core model also allows the user to exploit dynamic exchanges between multiple viewpoints (i.e classifications) on the database. The core model of the MicroNOMAD tool strongly derives from the multimap topographic model which has been successfully tested on textual data in the framework of the NOMAD IR System. This latter model can be itself considered as a extension of the basic Kohonen's topographic map model. The MicroNOMAD core model added-value is then mainly to develop a synergy between the browsing and discovering capabilities of the NOMAD's original multimap model, on the one hand, and the natural capability of the imbedded Kohonen map model to support at the same time concept mapping and image mapping, on the other hand.