Extensibility in a Visual Language for Web-based Interpersonal Communication

One of the challenges of developing a visual language for human-to-human communication is coming up with a large enough visual vocabulary that users are not overly limited in the meanings they can express. If a language system can provide extension mechanisms directly to its end users, then the language, or various dialects of it, can grow with their use. Using an experimental prototype of a visual language called Vedo-Vedi, we present a mechanism that permits its users to de ne new objects, new icons for those objects, and relationships among those objects. The de ned objects can be used to ll slots in frames that represent visual sentences, and they take places within a hierarchical menu system. One of the challenges of providing such an extension mechanism for a visual language is making it easy for users to dene new icons, even if they are not comfortable with artistic drawing tools. In Vedo-Vedi, users can create new icons by selecting arbitrary rectangular regions within special images provided for the purpose, or for that matter, within any of the more than 80 images used in the system. Other possible extension mechanisms are also discussed along with implementation and performance issues.