Language Modeling for Automatic Speech Recognition of Inflective Languages

The book "Language Modeling for Automatic Speech Recognition of Inflective Languages" offers a comprehensive analysis of language modeling and automatic speech recognition for inflective languages, such as Slavic languages. These languages pose a significant challenge for existing speech recognition systems, as they typically do not perform as well as English. The authors, Zdravko Kacic and Gregor Donaj, highlight the essential linguistic features that are important for the development of effective speech recognition systems. Through a detailed error analysis focusing on the morphological properties of the words in the recognized sentences, it demonstrates how language models can be developed and integrated into speech recognition systems. The book is aimed at professionals and researchers working with speech recognition in large vocabularies as well as with continuous and spontaneous speech.