Ontology-based Semantic Support to Improve Accessibility of Graphics

We aim to ease the process of authoring accessible graphics as well as taking a first step towards the long-term goal of allowing blind persons to access graphics autonomously. We are developing and experimenting with a hierarchical set of knowledge bases related to the presentation of visual objects and cues in the form of ontologies that will act as the formal, axiomatic underpinnings of an accessibility layer or, later on, a graphics reader/browser for blind and visually impaired people. The concept and prototypes of smart (or communicative) graphics [1], in which readers obtain information about the syntactic and semantic content through the use of e.g. a natural language interface, should be expanded by exploiting the benefits of formal semantics supported by domain- and task-aware ontologies describing the elements, visual cues and relations used for visualization or visual display.