Using the World Wide Web of today, searching for a graphic pertaining to a particular subject domain or in response to a specific query is a difficult task. A typical search for a graphic related to a specific subject matter or query may yield hundreds or thousands of Web resources, few of which relate to the intended meaning. The primary goal of the completed dissertation is to develop and assess the feasibility of using a global ontology for Scalable Vector Graphics (SVG) written in the Semantic Web markup language Web Ontology Language (OWL).
SVG is an eXtensible Markup Language (XML) based technology used to describe two-dimensional graphics. SVG has the ability to fully scale images without loss of resolution, provide file sizes that are independent of resolution, represent text as text strings allowing the graphic to be fully searched for content, and support a rich set of geometrical primitives.
An SVG OWL ontology provides three benefits. First, the ontology enables powerful semantic search engines to quickly and efficiently pinpoint SVG graphics and relate these graphics to specific knowledge domains. Second, the ontology enables semantic search engines to understand the content of a SVG graphic and infer relationships between the content of the graphic and specific domains. Lastly, enabling SVG graphics to be annotated in varying levels of abstraction allows the graphic to be reused in other contexts.
The research methods included developing the framework for the model, identifying the entities to be used in the ontology, representing the conceptual elements using Unified Modeling Language (UML), converting the UML to OWL, evaluating the ontology to ensure that it meets the requirements initially presented, developing a working system based on the ontology and testing this system, and documenting the development process.
Regarding experimental results, a total of 69 queries were applied to a set of 500 images representing a range of both primitive and derived spatial properties. Both recall and precision were perfect, indicating the feasibility of effective ontology-based search for annotated vector graphics through this approach. The question of scalability to more complex and realistic settings remains for future research.
| Identifer | oai:union.ndltd.org:nova.edu/oai:nsuworks.nova.edu:gscis_etd-1232 |
| Date | 01 January 2014 |
| Creators | Mathis, Regina Mitzie |
| Publisher | NSUWorks |
| Source Sets | Nova Southeastern University |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | CEC Theses and Dissertations |
Page generated in 0.0019 seconds