Programming SVG Advanced Access Using Metadata and Fuzzy Sets

SVG is an XML namespace which is used to represent graphics and animation. This representational capability allows GIS data / information to be mapped to SVG elements and thereby display visuals (both GIS and 'diagram') on screen or paper.

This paper and its related tutorial discusses how metadata and also Fuzzy Sets can augment SVG / GIS and SVG / diagrams. SVG 1.0 is designed with internal / embedded metadata capability principally through its "metadata" element. While RDF is (considered to be) the default metadata language, the SVG metadata element allows other metadata representations, such as XTM, DAML-OIL, and XGMML, in addition to RDF.

This author has shown, in previous papers, usage of mixed mode metadata, where a single SVG file contained both XML RDF and Topic Map metadata. This author's example of metadata in the SVG 1.0 Specification illustrates mixed mode metadata.

In SVG, metadata is text placed in the SVG metadata element. Metadata information may be placed in the SVG Title and Description elements also because of the namespace capability, however, this author does not believe this is a good practice.

Because the SVG parser does not attempt to use the metadata we are free to have any number of external metadata processors. This makes it practical and convenient to augment (use of) RDF with metadata representations that are 'kinder'. This paper shows use of RDF(S) and XTM. RDF metadata is created to show what it is and what it can represent for SVG. (including illustration of W3C's 'Linearization' approach.) This paper then goes on to show how RDF predicates can be used as locators of the facts they describe. SVG pictures are shown "parsed" for content using RDF facts, and fuzzy set parameterization. XTM topic maps are shown to display what they are and what they can represent for SVG, which is quite a bit. They add a dimension of searchability/access not available from RDF.

Fuzzy Sets are a branch of mathematics pioneered by Lotfi Zadeh at Berkeley University California. Many papers have been written about the mathematical details as to how fuzzy sets are computed and applied to various representational problems, by Zadeh, Mamdani, Sugeno and others. This author had a paper on fuzzy set application recommended for publication in an international technical journal. The paper was how to use fuzzy sets to describe spatial constructs using situatedness. In this paper we see using fuzzy sets for adding non-binary spatial and temporal constructs to SVG descriptions, permitting continuous accessibility.

Fuzzy Sets are a bone fide way of consistantly handling problems caused by First Order Predicate Logic's "Law of the Excluded Middle" (whereby the values can only be either 0 or 1 and nothing exists in between.) Of course the real world is continuous not binary. This paper shows use RDF(S) representation of spatiality and SVG object relationship and how application of Zadeh's fuzzy set mathematics overcomes RDF's everything is "black or white".

An SVG diagram (a "business chart" and an electronic circuit schematic) and an SVG animation are presented and the code is shown / explained which adds metadata to these visuals and also the practical use of fuzzy sets, which is the parameterization of RDF's predicates, and to provide a continuous measure rather than RDF's binary one. XGMML is shown as a network representation of 'significant' visual information, which is searchable and castable to RDF or other metadata representations.

The methodology for serializing ascription of SVG visual content (ie making natural language text from the picture itself) is shown (including 'non-linear' contexts). Each of the sections of this paper have accompanying code shown during the presentation.

Creating some fuzzy sets and displaying them with SVG is shown and explained. The SVG display allows for immediate understanding of how fuzzy sets augment the accessability of SVG visual content. Fuzzy set information can also be serialized and output as text or text-to-speech voice.

We see the creating of XGMML XML structures from SVG files. The paper also presents creating XGMML XML structures from RDF

Presented by David Dodds one of the original authors of SVG 1.0