Advances in Machine Translation for Low-Resource Languages and Domains

Dear Colleagues,

In the context of a globalizing world, accurate Machine Translation (MT) systems are becoming indispensable engines for breaking the language barriers between countries. MT systems alongside the Internet are becoming major solutions that companies will rely on for promoting their products across borders, enabling them to get in touch with customers and understand their feedback (sentiment, opinion, etc.) regardless of their native language. However, building an accurate MT system for any pair of languages requires substantial resources and knowledge for modelling both languages. For instance, many years of expert work are required to add a new language pair to a rule-based MT system, which is, despite the long history of this approach, one reason why at present only a very limited number of language pairs are covered, and these tend to comprise only the most common languages. On the other hand, there are Statistical MT (SMT) systems (of which the example-based systems are a variant) and Neural MT (NMT) systems, which try to learn how to translate by analyzing the translation patterns found in large collections of human translations. As the statistical and neural algorithms used in these systems are largely language-independent, they can be quickly adapted to new language pairs. The amount of research that has been devoted to Statistical MT and Neural MT has led to some important achievements and improvements for certain pairs of languages. However, the current state of MT systems for low-resource languages and domains has not reached the required quality in order to be used at a large scale. Indeed, the creation of MT systems is more complex as (1) the usage and meanings of words are adapted and modified in the language of specialized domains and genres, and (2) languages evolve over time—new topics and disciplines require the creation or borrowing (e.g., from English) of new terms, with other terms becoming obsolete. In addition, statistical MT and neural MT do not work well for morphologically rich languages, unless the amount of training data is very large.
The objective of this Special Issue is to promote research and discussion, as well as reflect on, the latest advances and findings especially related to the use of advanced deep neural models to address neural machine translation issues. This Special Issue welcomes researchers and practitioners from industry and academia to contribute original research work developed using recent technologies such as Deep Learning and Artificial Intelligence to handle machine translation for low-resource languages and domains.

Topics include but are not limited to:

• General research on Machine Translation (MT).
• Transfer-learning techniques for low-resourced languages MT (use of multilingual, pre-trained models, unsupervised, semi-supervised, zero-shot, few-shot training, etc.).
• MT for morphologically rich languages.
• MT for low resource languages.
• MT for specialized domains.
• Measuring MT quality.
• Taking multiword expressions into account in MT.
• Semantics-based MT.
• Real time MT
• Hybrid approaches for MT

Dr. Nasredine Semmar
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Information is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.