Enhanced Techniques to Improve Speech Query based Tamil to English Cross Language Text Retrieval System using Ensemble Models

Abstract

Major objectives To design and develop a Tamil Speech Query Recognition system that uses waveletbased methods for noise removal and enhanced feature extraction optimized through the use of SelfOrganizing Map and ensemble Support Vector Machine SVM classifier for speech recognition To design and develop optimizing methods to convert the recognized Tamil text newlineinto English text To design and develop Heterogeneous Ensemble Model that uses hybrid associative and improved KNN algorithm for document retrieval