Technical Reports

The report FIMU-RS-2008-12

Verbalizing Visual Data for the Blind: Towards a More Complex Graphical Ontology

by Petr Peòáz, December 2008, 18 pages.

FIMU-RS-2008-12. Available as Postscript, PDF.


The standard situational and picture descriptions, as produced by human professionals, are being compared to the existing proposals of a dialogue-based processing and graphical ontologies. The ontology proposed to build up verbal presentations of the graphics for sightless persons seems to be very suitable for unambiguous cases. In order to include more complex cases of the visual communication, the existing ontology should be extended. Constituent elements of the graphical ontology in a broader sense are:

  • user`s perspective (user`s individual features, where the user is coming from to observe the picture, what he was doing before and what he is doing now while observing it),
  • intentional perspective of the author of the picture (picture encoding key),
  • functional perspective concerning the graphics (framework encoding key, relationship between the pictures in the same document, or in other documents related).
As to the third element, it is being demonstrated that the verbal expression of a picture which has been included in a document:
  • remains partially unchanged even in the new context,
  • needs reassessment, as to the rest (due to new ties and interpretation).
To display both information packages correctly to the user, the ontology must be complex enough to differentiate between formally identical questions and to answer them in a specific way according to the above-mentioned perspectives.

Responsible contact: unix(atsign)fi(dot)muni(dot)cz