Лінгвіcтична параметризація текстів для тренування користувачів у системі автоматичного розпізнавання англійського мовлення Dragon Naturally Speaking

Abstract

The report deals with the complex analysis of the training session texts taken from Dragon Naturally Speaking program for automatic speech recognition. The new type of software - SEGMIX© is introduced in order to provide automatic text segmentation into words with statistics frequency calculations attached. The results of the frequency calculations are analyzed from the point of view of artificial text vocabulary simplification which helps to optimize man-machine interaction.