ISO 24627-3:2021 pdf free download – Language resource management一 Comprehensive Annotation Framework (ComAF)一 Part 3: Diagrammatic semantic authoring (DSA).
Graphs (diagrams consisting of nodes and links) have been used for decades to represent and visualize both documents (instance data) and data schemas. This document concerns graph-based representation (not visualization) of documents (not data schemas).
Graph-based representation and visualization of documents are addressed by concept maps,[’l mind maps, argument maps, and so on. Theoretical linguistics and artificial intelligence have also used graph-based content visualization associated with semantic network, mental space,W] discourse representation structure,[Ul and so forth.
Graph-based visualization of data schemas (or ontologies, terminologies, metamodels, etc.) is a more usual practice. Ontologies are often visualized as graphs in which nodes are classes (and datatypes) and links are properties (relations). ISO 2415 6-1 specifies a UML-based visualization of concept modelling. Other metamodels are usually represented as similar diagrams, too.
This document gives a data schema of graph documents to facilitate composition and comprehension by making logical document structure explicit. It neither covers visualizations or manipulations of graphs nor does it define annotations to existing documents, but rather it addresses graphical/diagrammatic representation of documents for the sake of semantic authoring: i.e., for people to directly view and manipulate syntactic/semantic structures on computer displays or their future alternatives. The linearity of traditional text documents is due to the linearity of speech languages, which constrains the interaction between people and documents, making it hard for people to read and write. DSA defines graphical/diagrammatic documents with more explicit structures than in text in order to make it easier for people to read and write. Documents based on DSA, together with some user interfaces involving appropriate visualizations and easy operations, can enhance collaborations among people and between people and machines.
DSA mainly deals with syntactic or document structures. It addresses some fragmentary semantic structures as well, but more systematic semantics (formal mapping between documents and their meanings or logical forms) can be provided by another specification so that machines better ‘understand’ DSA-based documents and thereby better assist information sharing and consensus building among people.
Figure 1 shows a workflow involving DSA and other types of documents. The DSA-based documents in the upper half can be automatically converted (while preserving propositional content) to and from machine-understandable documents based on appropriate standards on semantic representations and annotations, It is possible to automatically generate traditional text documents from these machine-understandable documents (while preserving the propositional content, too), though the inverse conversion cannot generally be automated. Since DSA-based documents (together with some appropriate user interfaces) are easier for people to compose and interpret than text documents, people can usually touch and see DSA-based documents whereas traditional documents could be used for legacy procedures (such as patent applications) and oral presentations.