Context Representation Used in Speech Recognition

In this paper we show how speech recognition can be enhance by combining context representation with traditional Digital Signal Processing (DSP) methods. We implement a hybrid system which uses a rule-based expert system to create a Conceptual Dependency (CD) representation of spoken input. Conceptual Dependencies are used for natural language understanding of written text, but until now have not been applied to speech recognition. Our hybrid systems uses a three-step process. First, we implement a speech recognition system based on Vector Quantization (VQ) techniques. Then, the recognized sentences are used to create CD representations. Finally, the CD representations are combined with the context surrounding them to reinterpret the speech input and increase the speech recognition accuracy. The context is represented using Scripts, which describe a possible sequences of events that an actor or entity may perform under certain conditions. Using context representation reduces the speech..