Application Oriented Language Modeling

Abstract

In the previous chapters we described the use of language models in speech recognition, morphosyntactic description (MSD) tagging, error rates and error analysis based on information from MSD tags. In this we will present aspects of speech recognition applications that can be related to those factors. We will first describe general consideration on speech recognition performance and whether they might or might not be important in a specific application. In the following sections we will then describe examples of more specific application and the factors to consider when measuring their performance. We will also present a procedure for language modeling for a specific application. In the last sections we will briefly describe some possible final applications in domains where errors with different lexical properties have a different impact on the recognition performance.