Dynamical language prediction for Korean language disorders in restricted domain

We developed an augmentative communication system that generates a well-formed sentence in Korean language by symbols for people with communication disorders. The system should place a minimal burden on the users. The augmentative communication aids combine the benefits of pre-stored messages with linguistic prediction capabilities to create unique acceleration techniques. Different from traditional approaches, the lexicon is organized along dimensions using a dynamic approach as a connectionist model. When the users press symbols, the software predicts verbs, which are displayed in scanning modes based on semantic markers. In order to generate the well-formed sentences, adequate postpositions follow it according to verbs for each input word