This book constitutes the referred proceedings of the 4th Conference of the Association for Machine Translation in the Americas, AMTA 2000, held in Cuernavaca, Mexico in October 2000. The 18 technical papers, seven system descriptions, and three user studies presented were carefully reviewed and selected for inclusion in the book. All current issues of machine translations, ranging from theoretical and foundational issues to systems evaluation and case studies, are addressed.
This book focuses on a relatively new branch of linguistics, i.e. machine translation. This field seems to be particularly interesting since computers continuously increase their role in our lives. The work presents the role of computers as automated translators, from the very first ideas of machine translation to the contemporary and commercially successful systems. Basic notions and concepts of machine translation, regarding the division of MT systems and the application of automatic translation are briefly explained in this work, along with some of the misconceptions about machine translations and the perspectives for the future development. The most important aspect of this study is the development of a simple experimental rule-based system based on a free Apertium engine. The process of creation of the system is presented in detail and the output of the system is confronted with some popular professional systems.
The native languages all over the world are growing rapidly along with the growth of technology, in general, and information technology, in particular. On the one hand, the world experiences a growth in the native language and on the other hand, precious and nascent information come through foreign languages. The demand for making the information available in native languages is increasing. Therefore, we need an efficient and practical method to fulfill this demand. This book describes one such method; English to Tamil Machine Translation System. Though the book elucidates English to Tamil translation system, the methodology can be extended to other language pair especially Indian Languages. This book is especially useful to the researchers in the field of Machine Translation and the student of Computational Linguistics course.
In the era of globalisation Machine Translation (MT) has become an important facility for international communication. This work aims to present a brief historical overview of MT, its evaluation and the analysis of the Machine Translation output. Chapter One focuses on the major events, ideas and approaches in the history of Machine Translation whose influences are visible in the field nowadays. Additionally, it investigates the state of the art in the field. Chapter Two is an attempt to present the human and automatic methods of evaluation of Machine Translation systems and engines and describes strong and weak points of each method. Chapter Three is devoted to the author’s analysis of the output of two English-Polish MT systems namely Techland’s English Translator XT and Google’s Google Translate.
This book is for the geeks who want to develop computers systems to translate human language. As the title suggests, in this book we discussed about Example-Based approach of Machine Translation for English-Bengali language pair. The author wrote this book from his industry level machine translation system development experience.This book can be useful for students to get a brief idea about the novel research area of Machine Translation. In this book we discussed about a real-life application development for English to Bengali machine translation. Researchers and developers interested to make machine translation system for their own language can get some brief guideline. Readers can feel free to contact the author for any guideline. The developed machine translation system will be freely available from www.akkhorbangla.com.
Learning languages is considered as one of the crucial issues nowadays. This fact is true for a number of reasons. On top of it, to eradicate the conflict possibly occurred due to poor communication between nations. Another reason is to strengthen global relations via understanding each other cultures. Finally, to enhance external trade between countries. Another important factor to be taken into account is translation. Through the translation process, cultures, documents, catalogs and etc.. can be translated into different languages for people to read and understand. Finally, performing machine translation (MT) can benefit not only a businessman but also a buyer and anyone seeks for information and knowledge. This is because MT saves money, time and effort. However, an MT can not takeover the human translator's own job as the machine can not produce a fully automatic high quality translation. Therefore, it is still in need to a human interaction either before, during or after the translation process. This book outlines some definitions, advantages and the future of machine translation. It also presents a comparison between human and machine translation out put.
In the era of Multidisciplinary Research, Computational Linguistics is a fascinating as well as challenging field of study. When Language is processed through the computational programming, especially in Machine Translation (MT), an exhaustive linguistic knowledge of both Source Language (SL) and Target Language (TL) is required for correct translation of the given input sentence. As a sentence has two primary divisions, Noun Phrase (NP) and Verb Phrase (VP), there are some challenging issues in its mapping. In this book, the author has discussed the challenges in NP Mapping for Sanskrit-Hindi Machine Translation (SHMT). The challenges in case mapping, adjective mapping indeclinable mapping, conjunct mapping, number and gender mapping as well as word order mapping needs Transfer Grammar Rules based on minute study of syntacto-semantic relationship between used padas.
Regression based machine translation (RegMT) model provides a learning framework for machine translation, separating learning models for training, training instance selection, feature representation, and decoding. Transductive learning approach employs training instance selection algorithms that not only make RegMT computationally more scalable but also improve the performance of standard statistical machine translation (SMT) systems. Sparse regression models for SMT are introduced and the obtained results demonstrate that sparse regression models perform better than other learning models in predicting target features, estimating word alignments, creating phrase tables, and generating translation outputs. We develop good evaluation techniques for measuring the performance of the RegMT model and the quality of the translations. We demonstrate that sparse L1 regularized regression performs better than L2 regularized regression in the German-English translation task and in the Spanish-English translation task when using small sized training sets. Graph based decoding can provide an alternative to phrase-based decoding in translation domains having low vocabulary.
Bangla (can also be termed as Bengali) is one of the richest languages among all the languages exist in the world and one of the largely spoken languages in the world. More than 220 million people speak in Bangla as their native language. It is ranked sixth based on the number of speakers. Bangla is the mother language of Bangladesh and also a large number of people in eastern area of India (West Bengal and Kolkata as its capital) speak and write in Bangla. This book presents the adapting rule based machine translation from English to Bangla. The proposed language translation model relies on rule based methodologies especially fuzzy rules. There are “If - Then” basis rules are applied for the English to Bangla language translation. In this language translation, we use rough set technique for Knowledge Representation System. This technique is used to classify each English sentence to a particular class using attributes of that English sentence and then translate them to the Bangla sentence using the rules that is produced earlier in the language translation system. Here English to Bangla bilingual dictionary has been formed for the purpose of language translation.
Machine translation research contains lots of howlers and misconceptions, but the classics of all time, and my personal favorite is as follows: Once an MT program translated “The spirit is willing, but the flesh is weak” into Russian and the back translation was “The vodka is good, but the meat is rotten”. This study is written to understand how far the quality has improved. It is expected that the study will contribute both to human translation and machine translation research by providing a comprehensive evaluation by human annotators and by providing an error categorization on the outputs of machine translation programs from English into Turkish.
News as a major part of today’s communications plays a considerable role in the exchange of information. This information exchange would be of value if it is proper and accurate. While there is not a united language for all people worldwide, there is no remedy of translation. In order to establish a proper communication and offer a meaningful translation, one needs to transfer the original utterances to his/her audiences thoroughly. It is possible when the translator is aware of source text (ST) and target text (TT) structure, text function in both ST and TT context, and culture-specific elements and the characteristics of the audiences. In other words, translation should be meaningful, accurate, and functional for its context. In this work, the author evaluated the quality of news translation in Iranian Students News Agency applying TQA model of Farahzad (1992), in which 5 criteria of appropriateness, accuracy, naturalness, cohesion and style were introduced as quality indexes. Thus the present work would be useful for translation instructors, translation studies students and teachers, and those who are concerned with media, language and quality of information exchange.
The monograph offers an insight into the problem of rendering emotions in literary artistic translation. The translation of interjections as synsemantic language units expressing emotions, volition or onomatopoeia is analysed in terms of translation equivalence, adequacy and fidelity. The methods of translating interjections in the English-Ukrainian and Ukrainian-English translation directions are analysed and assessed for applicability.
The present volume is the fruits of the BA projects in the School of Foreign Studies of Yangtze University with special regard to translation. It contains ten theses which deal with ten topics in Translation Studies, including culture in idioms and translation, translation teaching for undergraduates, tourism translation, kinship terms and translation, English syntax and translation, classical Chinese poetry and translation, Chinese syntax and translation, translation competence, news translation and advertising translation.
Teaching computers how to translate from one human language into another is an ongoing research topic in computer science. This book explains the fundamental ideas on which statistical machine translation is based and what you need to get a real statistical machine translation system running. The core of this book is focused on different strategies that will help you minimize the time needed to train your system and to get the best translation possible. We pose the question whether it is primarily the size of your training corpus that determines the quality of translation or if there are other ways that can boost the quality of the translated text as well. What is the best approach to deal with morphologically rich languages? How important is the genre of the text? We use tools like Moses, MGIZA, SRILM and perform empirical analyses to answer these questions.
The object of the work is the analysis of translation universals in the English and Spanish translations of the novel Saturnin by Zdenek Jirotka. The aim of the analysis is to find out how the two translations differ in respect to the number and distribution of translation universals and to try to identify the potential cause(s) of these differences. The Thesis is divided into two parts, the Theoretical and the Practical Part. The Theoretical Part provides the necessary background to the analysis, focusing on the basic theoretical issues in the field. In addition, information about the life and works of Zdenek Jirotka are provided, as well general information about the two translations and their translators. In the Practical Part, eight sections selected from the novel are analysed separately for the presence of translation universals. Each section focuses on one or two types of translation universals only. The analysis is concluded by the summary of the results in each translation and by the comparison of both translations.
This book is full of strategies to deal with loan terms in technical translation. Taking the reader through the translation universals and corpus-based translation, the author superbly discusses the strategies to deal with non-equivalence in translations, especially in legal texts. The author investigates lexical simplification as a translation universal and how it is accounted for in the English-to-French legal translation of Latinisms. Within descriptive and functional approaches to translation, the author discusses how loan terms are dealt with when they are accepted in the source language and not lexicalized in the target language. This book is indispensable for transtion students as well as practising translators. They will learn how English-to-French translations are system-specific, convention-specific, function-specific rather than translation-specific.
In the contemporary world, translation has an increasingly important role to play in communication across cultures and languages. Although translation has a central core of linguistic activity, it belongs more properly to semiotics. Intersemiotic translation involves translation between two different media, for example, from the verbal medium into the pictorial medium. This category made it possible for different sign systems to be examined through the prism of Translation Studies for book illustrations to be seen as a type of translation of the text in the illustrated book. The conversion of verbal information into other kinds of non-verbal information is a rather complex issue, involving not only the characterization of the laws governing the types of signs but the analysis of both media as source and target works, which they will imply. The text can be regarded as the primary source because it is usually the first work to be created, the pictures being derived from it. So, comic book or graphical story form is a complicated system which fundamentally makes from visual signs and linguistic-written signs. This book may, hopefully, suggest new insights to translation studies.
Machine translation is about an automatic conversion of one natural language text to another using computer. This book explains the development of Tamil Linguistic Tools and English-Tamil Machine Translation System. Development of Tamil linguistic tools such as Part-of-Speech Tagger, Morphological Analyzer and Morphological Generator are also explained in detail. Statistical Machine Translation (SMT) system treats translation problem as a machine learning problem. Learning algorithms create a model from parallel sentences and using this model, new sentences are translated. Huge amounts of parallel corpora are required for learning the pattern. SMT algorithms are difficult to apply directly for dissimilar language pair like English-Tamil. To resolve this and improve the translation performance an additional morphological pre-processing is required. This pre-processing is performed using Tamil linguistic tools which augment the linguistic information into the parallel corpora. The pre and post processing methodology explained in this book are applicable to other language pairs too.
The aim of this research is to develop a method to automatically build translation memories to be used in the subtitling task. The inputs to create the corpus are the novels the audiovisual material is based on. The methodology includes the automatic identification and extraction of dialogues, and their alignment for the purpose of building a translation memory for subtitling. The aligned bi-texts will serve as translation memories to be employed in the translation of films based on these novels.
For readers in the English-speaking world, almost all Holocaust writing is translated writing. Translation is indispensable for our understanding of the Holocaust because there is a need to tell others what happened in a way that makes events and experiences accessible – if not, perhaps, comprehensible – to other communities. Yet what this means is only beginning to be explored by Translation Studies scholars. This book aims to bring together the insights of Translation Studies and Holocaust Studies in order to show what a critical understanding of translation in practice and context can contribute to our knowledge of the legacy of the Holocaust. The role translation plays is not just as a facilitator of a semi-transparent transfer of information. Holocaust writing involves questions about language, truth and ethics, and a theoretically informed understanding of translation adds to these questions by drawing attention to processes of mediation and reception in cultural and historical context. It is important to examine how writing by Holocaust victims, which is closely tied to a specific language and reflects on the relationship between language, experience and thought, can (or cannot) be translated. This volume brings the disciplines of Holocaust and Translation Studies into an encounter with each other in order to explore the effects of translation on Holocaust writing. The individual pieces by Holocaust scholars explore general, theoretical questions and individual case studies, and are accompanied by commentaries by translation scholars.