Ontology-based Semantic Support to Improve Accessibility of Graphics

Area of Science:

• Computer Science
• Human-Computer Interaction
• Artificial Intelligence

Background:

• Current methods for creating accessible graphics are cumbersome.
• Blind and visually impaired individuals face significant barriers in accessing visual information.
• Existing 'smart graphics' concepts can be enhanced with formal semantics.

Purpose of the Study:

• To simplify the authoring of accessible graphics.
• To develop foundational knowledge bases (ontologies) for an accessibility layer.
• To enable autonomous graphics access for blind and visually impaired users.

Main Methods:

• Developing and experimenting with hierarchical ontologies.
• Defining formal semantics for visual objects, cues, and relations.
• Integrating domain- and task-aware ontologies into graphics presentation.

Main Results:

• Prototypes of smart graphics incorporating formal semantics are being developed.
• Ontologies provide a formal, axiomatic basis for an accessibility layer.
• The approach aims to enhance information retrieval from graphics via natural language interfaces.

Conclusions:

• Ontologies are crucial for creating an accessibility layer for graphics.
• This work is a step towards autonomous graphics interpretation for visually impaired users.
• Formal semantics in graphics enhance understanding of visual content and relations.