Assessing the support provided by a toolkit for rapid prototyping of multimodal systems

Abstract

Choosing an appropriate toolkit for creating a multimodal interface is a cumbersome task. Several specialized toolkits include fusion and fission engines that allow developers to combine and decompose modalities to capture multimodal input and provide multimodal output. Unfortunately, the extent to which these toolkits can facilitate the creation of a multimodal interface is hard or impossible to estimate, due to the absence of a scale where the toolkit's capabilities can be measured on. In this paper, we propose a measurement scale, which allows the assessment of specialized toolkits without need for time-consuming testing or source code analysis. This scale is used to measure and compare the capabilities of three toolkits: CoGenIVE, HephaisTK and ICon.