From text description to animation sequences

Most current user manuals for electro-mechanical devices are text-based. The inclusion of interactive 3D graphics and multimedia adds physical dimension and realism to “reading” the manual, and encourages the users to explore and experiment with the manual. The new manual should provide the seamless environment for the users to browse the manual in either text or graphics modes and to navigate freely between the two modes. This paper describes the development of a 3D graphical manual that supports the automatic translation of text descriptions in manual pages to animation. The system uses the frame formalism to model the relationships and dependencies between the graphical components of the device. A concept-based information retrieval (IR) engine is used to perform the text-to-animation translation. In order to eliminate the need to construct a separate concept structure for the IR engine, the system uses the same frame-based graphics model for both the animation and text-to-animation translation. The feasibility of the system is demonstrated using an electronic manual for a laser printer