TAILIEUCHUNG - Báo cáo khoa học: "Does more data always yield better translations?"

Nowadays, there are large amounts of data available to train statistical machine translation systems. However, it is not clear whether all the training data actually help or not. A system trained on a subset of such huge bilingual corpora might outperform the use of all the bilingual data. This paper studies such issues by analysing two training data selection techniques: one based on approximating the probability of an indomain corpus; and another based on infrequent n-gram occurrence. Experimental results not only report significant improvements over random sentence selection but also an improvement over a system trained with the whole. | Does more data always yield better translations Guillem Gasco Martha-Alicia Rocha German Sanchis-Trilles Jesus Andres-Ferrer and Francisco Casacuberta Departament de Sistemes Informatics i Computacio Universitat Politecnica de Valencia Cami de Vera s n 46022 Valencia Spain ggasco mrocha gsanchis jandres fcn @ Abstract Nowadays there are large amounts of data available to train statistical machine translation systems. However it is not clear whether all the training data actually help or not. A system trained on a subset of such huge bilingual corpora might outperform the use of all the bilingual data. This paper studies such issues by analysing two training data selection techniques one based on approximating the probability of an indomain corpus and another based on infrequent n-gram occurrence. Experimental results not only report significant improvements over random sentence selection but also an improvement over a system trained with the whole available data. Surprisingly the improvements are obtained with just a small fraction of the data that accounts for less than of the sentences. Afterwards we show that a much larger room for improvement exists although this is done under non-realistic conditions. 1 Introduction Globalisation and the popularisation of the Internet have lead to a rapid increase in the amount of bilingual corpora available. Entities such as the European Union the United Nations and other multinational organisations need to translate all the documentation they generate. Such translations happen every day and provide very large multilingual corpora which are oftentimes difficult to process and significantly increase the computational requirements needed to train statistical machine translation SMT systems. For instance the corpora made available for recent machine translation evaluations are in the order of 1 billion running words Callison-Burch et al. 2010 . However two main problems arise when attempting to use this huge pool

• Báo cáo khoa học
• Does more data
• yield better translations
• Guillem Gasc
• scientific reports
• model language
• process natural language
• Dialog Control
• Natural Language System
• Dialog Control in a Natural Language System
• Expanding the Horizons
• Natural Language Interfaces
• Phil Hayes
• PROSPECTS OR PRACTICAL
• NATURAL LANGUAGE SYSTEMS
• Workshops
• NATURAL LANGUAGE
• INTERACTION WITH MACHINES
• Long Papers
• COMPUTER INTEBFACE DESIGN MURRAY TUROFF DEPARTMENT OF COMPUTER
• IiVFORMATION SCIENCE IIEW JERSEY INSTITUTE OF TECHNOLOGY
• EVALUATION OF NATURAL LANGUAGE
• INTERFACES TO DATABASE SYSTEMS
• A PANEL DISCUSSION
• database query system
• J
• Norwood Crout
• INTERFACES TO DATA BASE SYSTEMS
• Bozena Henisz
• INTERPRETING NATURAL LANGUAGE
• DATABASE UPDATES
• S
• Jermld Kaplan Jim David
• TRANSPORTABLE NATURAL LANGUAGE
• INTERFACES TO DATABASES
• Gary G
• Hendrlx and William H
• Lewis
• DATABASES THEORETICAL
• TECHNICAL ISSUES
• TRANSPORTABLE NATURAL LANGUAGE INTERFACES
• PROBLEMS AND TECHNIQUES
• Barbara J
• Grosz
• THEORETICAL TECHNICAL ISSUES
• NATURAL LANGUAGE ACCESS TO DATABASES
• R
• Petrick
• Domain Independent
• Database Access Systems
• ISSUES IN NATURAL LANGUAGE
• ACCESS TO DATABASES
• A LOGIC PROGRAMMING PERSPECTIVE
• Sharon C
• Salveter David Maier
• PLANNING NATURAL LANGUAGE
• EXPRESSIONS REFERRING
• Douglas E
• Appelt
• THE TEXT SYSTEM
• NATURAL LANGUAGE GENERATION
• Kathleen R
• M
• MENTING A DATABASE
• KNOWLEDGE REPRESENTATION
• THE KEY TO THE SELECTION PROBLEM
• E
• Jeffrey Conklin David D
• McDonald
• A KNOWLEDGE ENGINEERING
• NATURAL LANGUAGE UNDERSTANDING
• Stuart C
• Shapiro
• NATURAL LANGUAGE TEXTS
• NECESSARILY GRAMMATICAL
• EVEN COMPLETE
• UNGRAMHATICALITY
• EXTRA GRAMMATICALITY
• NATURAL LANGUAGE UNDERSTANDING SYSTEMS
• Taxonomy
• Descriptions
• Individuals in Natural
• Language Understanding
• The Use of Ooject Special Knowledge
• Natural Language Processing
• Mark H
• Bursteln