Latent Semantic Analysis and Machine Translation

Abstract

Computer-based translation systems are not rivals to human translators, but they are aids to enable them to increase productivity in technical translation. Machine translation aims to undertake the whole translation process, but whose input must invariably be revised. Latent Semantic Analysis and Probabilistic Latent Semantic Analysis are two newly developed computational models which their application in machine translation will solve some of the problems facing machines in accounting for the way human knowledge is comprehended.