This item is non-discoverable
Development of Machine Translation Models: A Systematic Review
Date
2020
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Advanced Natural Language Processing (ANLP) has a wide variety of domains, including
machine translation (MT) and its numerous models. It specifically focuses on two models
of MT, which are the Statistical Machine Translation (SMT) model and Neural Machine
Translation (NMT) model. SMT operates by using a database of existing information and
the probability distribution of the original and target languages in order to perform
translation. By contrast, NMT uses deep and representative learning in order to perform
the same task. Some of the major challenges faced by machine translation technologies
are posed by the dynamic fluidity of human language and the major contrast among
different languages. Due to this, the MT industry experienced a major development since
its establishment. This research paper aims to explore in-depth the field of MT and find
out the latest developments in this area as well as to compare and contrast the different
translation models: namely, SMT and NMT. A systematic literature review has been
conducted in order to find highly reputed peer-reviewed papers investigating the same
topic. A set of research questions has been developed and their rationale has been
explained. The research strategy used to conduct this study is based on a thorough search
on academic databases using keywords derived from the research questions. Finally, pre selection and selection criteria have been examined and cross-referenced in order to find
the most important pieces of literature. Based on the literature review, the research
questions have been addressed. Besides, we highlighted MT methods, which aim to
improve the quality of the translations that they produce.