作者: Michael Johnston
DOI:
关键词:
摘要: Multimodal interfaces enable more natural and effective humancomputer interaction by providing multiple channels through which input or output may pass. In order to realize their full potential, they need support not just from modes, but synchronized integration of semantic content different modes. This paper describes a multimodal language processing architecture allows for declarative statement strategies in unification-based grammar formalism. The is currently deployed working system enabling with dynamic maps using speech pen, the approach general supports wide variety other potential interfaces.