Voice Navigation

Model

Data Required for voice enabling links

• Anchor Text => Corresponding Grammar Rule
• If it is a Link or Script ?
• Link Location or Script Code

Data Required for creating general links (not necessarily voice enabled)

• Anchor Text => Corresponding Grammar Rule
• If it is a Link or Script ?
• Link Location(code for href attr)
• Script Code (code for onclick)
• Other attributes for "A" html tag.
• Other code besides anchor text to be inserted between opening and closing "A" tag(eg:- Image).

A createlink function in weblib will allow the user to the following:

• Create a General Link
• Create a voice enabled link

Issues Considered

• Anchor Text rules resolved to some extent. This solves most cases[say 95% or more] but is not perfect.

Open Issues

• Resolving anchor text names within a page so that resulting voice rules do not conflict. Also device a strategy make anchor texts phonetically different.
• Analyse if it is required to make links which are not visible to human eye on the page ( in span and div ) ie which get visible on calling scripts should not be made voice enabled.
• A computation intensive algorithm should not be devised as it will increase the loading time.