The goal of this project is to improve incorporation of context in processing streams of audio and/or text. This work is being done within Pliers to improve features in Neuroscout and also in consideration of spoken or text based dialogue systems.