Reference Resolution as a facilitating process towards robust Multimodal Dialogue Management : A Cognitive Grammar Approach

This paper tries to fit a novel reference resolution mechanism into a multimodal dialogue system framework. Essentially, our aim is to show that a typical multimodal dialogue system can actually benefit from the cognitive grammar approach that we adopt for reference resolution. The central idea is to construct and update reference and context models in a manner that imparts adequate level of under-specificity to multimodal semantics. Context-independent semantic representations are constructed based upon the surface structure of the referring expressions and syntactic constraints within an utterance. The reference resolution algorithm assimilates these semantic representations into a coherent context model, resulting in the profiling of the intended referent. The resolution model is built upon discursive, perceptual and conceptual cues, thus successfully accounting for multiple modalities and a multi-dimensional application domain model.