Generating Multimodal Grammar using CYK and e-GRIDS Algorithm?

To communicate with each other human being uses multiple modes like text, speech, hand gesture, facial expressions. The use of theses modes makes human communication more fast and flexible. In previous years several methods are used to bring human computer communication more closely. It makes high costs for developing and maintaining multimodal grammar. Integrating and understanding multiple input interfaces leads to discover the novel algorithm which provides a way to automate the grammar and update the grammar. In this paper, we propose a grammar algorithm that uses multimodal grammar from positive samples of multimodal sentences from multiple inputs. The first phase of the algorithm generates multimodal grammar which is able to parse positive samples of sentences collected from multiple inputs and second phase updates grammar by using learning operators and minimum length metrics for avoiding over generalization problem.