Academia.eduAcademia.edu

Outline

Title
Abstract
Introduction and Background
The MT Methodology in Brief
Objective Evaluation Experiments
Subjective Evaluation Results
Discussion
References
All Topics
Computer Science
Natural Language Processing

Language-independent hybrid MT with PRESEMT

Profile image of Marina VassiliouMarina Vassiliou

Abstract

The present article provides a comprehensive review of the work carried out on developing PRESEMT, a hybrid language-independent machine translation (MT) methodology. This methodology has been designed to facilitate rapid creation of MT systems for unconstrained language pairs, setting the lowest possible requirements on specialised resources and tools. Given the limited availability of resources for many languages, only a very small bilingual corpus is required, while language modelling is performed by sampling a large target language (TL) monolingual corpus. The article summarises implementation decisions, using the Greek-English language pair as a test case. Evaluation results are reported, for both objective and subjective metrics. Finally, main error sources are identified and directions are described to improve this hybrid MT methodology.

Related papers

Hybrid Machine Translation
Eva Hasler

2007

View PDFchevron_right
Wider context by using bilingual language models in machine translation
Stephan Vogel

2011

Abstract In past Evaluations for Machine Translation of European Languages, it could be shown that the translation performance of SMT systems can be increased by integrating a bilingual language model into a phrase-based SMT system. In the bilingual language model, target words with their aligned source words build the tokens of an n-gram based language model. We analyzed the effect of bilingual language models and show where they could help to better model the translation process.

View PDFchevron_right
Implementing a language-independent MT methodology
George Tambouratzis

The current paper presents a languageindependent methodology, which facilitates the creation of machine translation (MT) systems for various language pairs. This methodology is implemented in the PRESEMT hybrid MT system. PRESEMT has the lowest possible requirements on specialised resources and tools, given that for many languages (especially less widely used ones) only limited linguistic resources are available. In PRESEMT, the main translation process comprises two phases. The first one, Structure selection, determines the overall structure of a target language (TL) sentence, drawing on syntactic information from a small bilingual corpus. The second phase, Translation equivalent selection, relies on models extracted solely from monolingual corpora to implement translation disambiguation, determine intra-phrase word order and handle functional words. This paper proposes extracting information for disambiguation from the monolingual corpus. Experimental results indicate that such information substantially contributes in improving translation quality.

View PDFchevron_right
EU-BRIDGE MT: Combined Machine Translation
Stephan Peitz

Proceedings of the Ninth Workshop on Statistical Machine Translation, 2014

This paper describes one of the collaborative efforts within EU-BRIDGE to further advance the state of the art in machine translation between two European language pairs, German→English and English→German. Three research institutes involved in the EU-BRIDGE project combined their individual machine translation systems and participated with a joint setup in the shared translation task of the evaluation campaign at the ACL 2014 Eighth Workshop on Statistical Machine Translation (WMT 2014).

View PDFchevron_right
Addressing Challenges in Multilingual Machine Translation
Rekha Sugandhi

The machine translation process may be unidirectional or bidirectional between a pair of languages. Or it can be multilingual too. A number of software's are developed till date and different advancements are taking place in this field to overcome the language barriers and create borderless marketplace. Still there are many challenges involved in this field of AI which are yet to be overcome. The translation quality of MT systems may be improved by developing better methods as well as by imposing certain restrictions on the input. All sort of challenges peak in case of Multilingual Machine Translation as compared to bilingual one. Paper focuses on long term challenges like High-Quality MT for many more language pairs, training with limited data resources, robustness across domains, genres and language styles, Achieving human-level translation quality and fluency.

View PDFchevron_right
Getting Past the Language Gap: Innovations in Machine Translation
Rodolfo Delmonte

Mobile Speech and Advanced Natural Language Solutions, 2012

In this chapter, we will be reviewing state of the art machine translation systems, and will discuss innovative methods for machine translation, highlighting the most promising techniques and applications. Machine translation (MT) has benefited from a revitalization in the last 10 years or so, after a period of relatively slow activity. In 2005 the field received a jumpstart when a powerful complete experimental package for building MT systems from scratch became freely available as a result of the unified efforts of the MOSES international consortium. Around the same time, hierarchical methods had been introduced by Chinese researchers, which allowed the introduction and use of syntactic information in translation modeling. Furthermore, the advances in the related field of computational linguistics, making off-the-shelf taggers and parsers readily available, helped give MT an additional boost. Yet there is still more progress to be made. For example, MT will be enhanced greatly when both syntax and semantics are on board: this still presents a major challenge though many advanced research groups are currently pursuing ways to meet this challenge head-on. The next generation of MT will consist of a collection of hybrid systems. It also augurs well for the mobile environment, as we look forward to more advanced and improved technologies that enable the working of Speech-To-Speech machine translation on hand-held devices, i.e. speech recognition and speech synthesis. We review all of these developments and point out in the final section some of the most promising research avenues for the future of MT. JH (ibid.,[1][2][3][4][5] Since then the field of Machine Translation (MT) has dramatically changed. And in the past 3 years, the field of MT has become so huge that there is no chance of sufficiently reviewing the whole spectrum of activities, tools and resources related to the field. Therefore, I will restrict this discussion to the leading and most promising approaches. 1 The first section will be devoted to examining in detail what Statistical Machine Translation (SMT) can offer in terms of improvements to the state of the art. Then, I will dedicate section "Hybrid and Rule-Based MT Systems" to hybrid methods and systems. In section "Syntax Based Approaches: From Hierarchical to SBSMT", I will delve into syntactically based MT systems. Then section "Knowledge-Based MT Systems" will introduce knowledge, semantically-based systems. Section "Evaluation Methods and Tools" will comprise an overview of evaluation methods and section "MT for the Future" will briefly give an overview of what in my opinion may constitute the future of MT systems. This section will comprise a subsection dedicated to Speech-To-Speech MT; in another subsection promising national projects will also be reviewed. This followed by the last section in which I draw conclusions.

View PDFchevron_right
Hybrid machine translation architectures within and beyond the EuroMatrix project
Teresa Herrmann

Proceedings of the 12th annual conference of the European Association for Machine Translation (EAMT 2008), 2008

Abstract. This paper presents two hybrid architectures combining rulebased and statistical machine translation (RBMT and SMT) technology. In the first case, several existing MT engines are combined in a multiengine setup, and a decoder for SMT is used to select and combine the best expressions proposed by different engines. The other architecture uses lexical entries found using a combination of SMT technology with shallow linguistic processing, which are then included in a rule-based system. The first architecture has ...

View PDFchevron_right
MMT: New Open Source MT for the Translation Industry
Nicola Bertoldi

2017

MMT is a new open source machine translation software specifically addressing the needs of the translation industry. In this paper we describe its overall architecture and provide details about its major components. We report performance results on a multi-domain benchmark based on public data, on two translation directions, by comparing MMT against state-of-theart commercial and research phrase-based and neural MT systems.

View PDFchevron_right
LIUM's SMT machine translation systems for WMT 2011
Haithem Afli

Proceedings of the …, 2011

This paper describes the development of French-English and English-French statistical machine translation systems for the 2011 WMT shared task evaluation. Our main systems were standard phrase-based statistical systems based on the Moses decoder, trained on the provided data only, but we also performed initial experiments with hierarchical systems. Additional, new features this year include improved translation model adaptation using monolingual data, a continuous space language model and the treatment of unknown words.

View PDFchevron_right
The Karlsruhe Institute of Technology Translation Systems for the WMT 2014
Mohammed Mediani

2014

In this paper, we present the KIT systems participating in the Shared Translation Task translating between English↔German and English↔French. All translations are generated using phrase-based translation systems, using different kinds of word-based, part-ofspeech-based and cluster-based language models trained on the provided data. Additional models include bilingual language models, reordering models based on part-of-speech tags and syntactic parse trees, as well as a lexicalized reordering model. In order to make use of noisy web-crawled data, we apply filtering and data selection methods for language modeling. A discriminative word lexicon using source context information proved beneficial for all translation directions.

View PDFchevron_right

References (28)

  1. Paul E. Black. 2005. Dictionary of Algorithms and Data Structures. U.S. National Institute of Stan- dards and Technology (NIST).
  2. Chris Callison-Burch, Cameron Fordyce, Philipp Koehn, Christof Monz and Josh Schroeder. 2009. Further Meta-Evaluation of Machine Translation. Proceedings of the WMT-08 Workshop, Colom- bus, Ohio.
  3. Chris Callison-Burch, Philip Koehn, Christof Monz, Omar F. Zaidan. 2011. Findings of the 2011Workshop on Statistical Machine Translation. Proceedings of the 6 th Workshop on Statistical Machine Translation, Edinburgh, UK, pp. 22-64.
  4. Jaime Carbonell, Steve Klein, David Miller, Michael Steinbaum, Tomer Grassiany and Jochen Frey. 2006. Context-Based Machine Translation. Pro- ceedings of the 7 th AMTA Conference, Cam- bridge, MA, USA, pp. 19-28.
  5. Michael Carl, Maite Melero, Toni Badia, Vincent Vandeghinste, Peter Dirix, Ineke Schuurman, Stella Markantonatou, Sokratis Sofianopoulos, Marina Vassiliou and Olga Yannoutsou. 2008. METIS-II: Low Resources Machine Translation: Background, Implementation, Results and Poten- tials. Machine Translation, 22 (1-2):pp. 67-99.
  6. Helena M. Caseli, Maria das Graças V. Nunes and Mikel L. Forcada. 2008. Automatic Induction of Bilingual resources from aligned parallel corpora: Application to shallow-transfer machine transla- tion. Machine Translation, 20:pp. 227-245.
  7. Michael Denkowski and Alon Lavie. 2011. Meteor 1.3: Automatic Metric for Reliable Optimization and Evaluation of Machine Translation Systems. EMNLP 2011 Workshop on Statistical Machine Translation, Edinburgh, UK, pp. 85-91.
  8. Ioannis Dologlou, Stella Markantonatou, George Tambouratzis, Olga Yannoutsou, Athanasia Fourla and Nikos Ioannou. 2003. Using Monolingual Corpora for Statistical Machine Translation: The METIS System. Proceedings of the EAMT-CLAW 2003
  9. Workshop, Dublin, Ireland, pp. 61-68.
  10. Richard O. Duda, Peter E. Hart and David G. Scott. 2001. Pattern Classification (2 nd edition). Wiley Interscience, New York, U.S.A.
  11. David Gale and Lloyd S. Shapley. 1962. College Admissions and the Stability of Marriage. Ameri- can Mathematical Monthly, 69:pp. 9-14.
  12. John Hutchins. 2005. Example-Based Machine Translation: a Review and Commentary. Machine Translation, 19:pp. 197-211.
  13. Alexandre Klementiev, Ann Irvine, Chris Callison- Burch and David Yarowsky. 2012. Toward Statis- tical Machine Translation without Parallel Cor- pora. Proceedings of EACL2012, Avignon, France, 23-25 April, pp. 130-140.
  14. Philip Koehn. 2010. Statistical Machine Translation. Cambridge University Press, Cambridge.
  15. John Lafferty, Andrew McCallum and Fernando Pereira. 2001. Conditional Random Fields: Prob- abilistic Models for Segmenting and Labelling Se- quence Data. Proceedings of ICML 2011, Belle- vue, Washington, USA, pp. 282-289.
  16. Harry Mairson. 1992. The Stable Marriage Problem. The Brandeis Review, 12:1.
  17. Stella Markantonatou, Sokratis Sofianopoulos, Olga Giannoutsou and Marina Vassiliou. 2009. Hybrid Machine Translation for Low-and Middle-Den- sity Languages. Language Engineering for Lesser- Studied Languages, S. Nirenburg (ed.), IOS Press, pp. 243-274.
  18. NIST 2002. Automatic Evaluation of Machine Trans- lation Quality Using n-gram Co-occurrences Sta- tistics.
  19. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. BLEU: A Method for Auto- matic Evaluation of Machine Translation. Pro- ceedings of the 40 th ACL Meeting, Philadelphia, USA, pp. 311-318.
  20. Jan Pomikálek and Pavel Rychlý. 2008. Detecting co- derivative documents in large text collections. Proceedings of LREC2008, Marrakech, Morrocco, pp.1884-1887.
  21. Prokopis Prokopidis, Byron Georgantopoulos and Harris Papageorgiou. 2011. A suite of NLP tools for Greek. Proceedings of the 10 th ICGL Confer- ence, Komotini, Greece, pp. 373-383.
  22. Felipe Sanchez-Martinez and Mikel L. Forcada. 2009. Inferring Shallow-transfer Machine transla- tion Rules from Small Parallel Corpora. Journal of Artificial Intelligence Research, 34:pp. 605-635.
  23. Helmut Schmid. 1994. Probabilistic Part-of-Speech Tagging Using Decision Trees. Proceedings of In- ternational Conference on New Methods in Lan- guage Processing, Manchester, UK, pp. 44-49.
  24. Temple F. Smith and Michael S. Waterman. 1981. Identification of Common Molecular Subse- quences. Journal of Molecular Biology, 147:195- 197.
  25. Matthew Snover, Bonnie Dorr, Richard Schwartz, Linnea Micciulla and John Makhoul. 2006. A Study of Translation Edit Rate with Targeted Hu- man Annotation. Proceedings of the 7 th AMTA Conference, Cambridge, MA, USA, pp. 223-231.
  26. Sokratis Sofianopoulos, Marina Vassiliou and George Tambouratzis. 2012. Implementing a language- independent MT methodology. Proceedings of the 1 st Workshop on Multilingual Modeling (held within the ACL-2012 Conference), Jeju, Republic of Korea, pp.1-10.
  27. Jinsong Su, Hua Wu, Haifeng Wang, Yidong Chen, Xiaodong Shi, Huailin Dong and Qun Liu. 2011. Translation Model Adaptation for Statistical Ma- chine Translation with Monolingual Topic Infor- mation. Proceedings of the 50 th ACL Meeting, Jeju, Republic of Korea, pp. 459-468.
  28. George Tambouratzis, Fotini Simistira, Sokratis Sofi- anopoulos, Nikos Tsimboukakis and Marina Vas- siliou. 2011. A resource-light phrase scheme for language-portable MT. Proceedings of the 15 th EAMT Conference, Leuven, Belgium, pp. 185- 192.

Related papers

OpenMT : Open Source Machine Translation Using Hybrid Methods TIN 2006-15307-C 03
Xabier Saralegi

2009

The main goal of the OpenMT project is the development of open source machine translation architectures based on hybrid models and advanced syntactic–semantic processors. These architectures combine the three main Machine Translation (MT) frameworks, Rule-based (RBMT), Statistical (SMT) and Example–based (EBMT), into hybrid systems. Defined architectures and results will be open source, allow for a rapid development and adaptation of new advanced machine translation systems for other languages. The project deals with four different languages: English, Spanish, Catalan and Basque. The translation systems developed for all language pairs will be internally evaluated using a rich set of linguistic metrics, and in different international evaluation campaigns.

View PDFchevron_right
Evaluating the Translation Accuracy of a Novel Language-Independent MT Methodology
Marina Vassiliou

The current paper evaluates the performance of the PRESEMT methodology, which facilitates the creation of machine translation (MT) systems for different language pairs. This methodology aims to develop a hybrid MT system that extracts translation information from large, predominantly monolingual corpora, using pattern recognition techniques. PRESEMT has been designed to have the lowest possible requirements on specialised resources and tools, given that for many languages (especially less widely used ones) only limited linguistic resources are available. In PRESEMT, the main translation process is divided into two phases, the first determining the overall structure of a target language (TL) sentence, and the second disambiguating between alternative translations for words or phrases and establishing local word order. This paper describes the latest version of the system and evaluates its translation accuracy, while also benchmarking the PRESEMT performance by comparing it with other established MT systems using objective measures.

View PDFchevron_right
OpenMT: Open Source Machine Translation Using Hybrid Methods
N. Areta

The main goal of the OpenMT project is the development of open source machine translation architectures based on hybrid models and advanced syntactic-semantic processors. These architectures combine the three main Machine Translation (MT) frameworks, Rule-based (RBMT), Statistical (SMT) and Example-based (EBMT), into hybrid systems. Defined architectures and results will be open source, allow for a rapid development and adaptation of new advanced machine translation systems for other languages. The project deals with four different languages: English, Spanish, Catalan and Basque. The translation systems developed for all language pairs will be internally evaluated using a rich set of linguistic metrics, and in different international evaluation campaigns.

View PDFchevron_right
Evaluating the Impact of Using a Domain-specific Bilingual Lexicon on the Performance of a Hybrid Machine Translation Approach
Nasredine Semmar

2015

This paper describes an Example-Based Machine Translation prototype and presents an evaluation of the impact of using a domainspecific vocabulary on its performance. This prototype is based on a hybrid approach which needs only monolingual texts in the target language and consists to combine translation candidates returned by a cross-language search engine with translation hypotheses provided by a finite-state transducer. The results of this combination are evaluated against a statistical language model of the target language in order to obtain the n-best translations. To measure the performance of this hybrid approach, we achieved several experiments using corpora on two domains from the European Parliament proceedings (Europarl) and the European Medicines Agency documents (Emea). The obtained results show that the proposed approach outperforms the state-of-the-art Statistical Machine Translation system Moses when texts to translate are related to the specialized domain.

View PDFchevron_right
Hybrid data-driven models of machine translation
Declan Groves

Machine Translation, 2005

This paper presents an extended, harmonised account of our previous work on combining subsentential alignments from phrase-based statistical machine translation (SMT) and example-based MT (EBMT) systems to create novel hybrid data-driven systems capable of outperforming the baseline SMT and EBMT systems from which they were derived. In previous work, we demonstrated that while an EBMT system is capable of outperforming a phrase-based SMT (PBSMT) system constructed from freely available resources, a hybrid ‘example-based’ SMT system incorporating marker chunks and SMT subsentential alignments is capable of outperforming both baseline translation models for French–English translation. In this paper, we show that similar gains are to be had from constructing a hybrid ‘statistical’ EBMT system. Unlike the previous research, here we use the Europarl training and test sets, which are fast becoming the standard data in the field. On these data sets, while all hybrid ‘statistical’ EBMT variants still fall short of the quality achieved by the baseline PBSMT system, we show that adding the marker chunks to create a hybrid ‘example-based’ SMT system outperforms the two baseline systems from which it is derived. Furthermore, we provide further evidence in favour of hybrid systems by adding an SMT target-language model to the EBMT system, and demonstrate that this too has a positive effect on translation quality. We also show that many of the subsentential alignments derived from the Europarl corpus are created by either the PBSMT or the EBMT system, but not by both. In sum, therefore, despite the obvious convergence of the two paradigms, the crucial differences between SMT and EBMT contribute positively to the overall translation quality. The central thesis of this paper is that any researcher who continues to develop an MT system using either of these approaches will benefit further from integrating the advantages of the other model; dogged adherence to one approach will lead to inferior systems being developed.

View PDFchevron_right

Related topics

  • Computer Science
  • Artificial Intelligence
  • Natural Language Processing
  • Machine Translation
    • 580 California St., Suite 400
    San Francisco, CA, 94104
    © 2025 Academia. All rights reserved