Recovering coherent interpretations using semantic integration of partial parses

This paper describes a chunk-based parser/semantic analyzer used by a language learning model. The language learning model requires an analyzer that robustly responds to extragrammaticality, ungrammaticality and other problems associated with transcribed language. The analyzer produces globally coherent analyses by semantically integrating the partial parses. Each resulting semantically integrated analysis is ranked by its semantic compatibility using a novel metric called semantic density.