ExampleBased Machine Translation Chinese to English

1
Example-Based Machine Translation (Chinese to
English)
Yu-Hui Yeh Supervisor Dr. Xue Li
English is the most common used language in the
world followed by Chinese. There is always a
demand of reliable and fast translation between
these two languages. This project is designed to
assist both Chinese and English background people
in exchanging information in these two languages,
especially in the way of browsing Chinese web
sites that are translated into the English
language. Current Research on Machine
Translation Machine Translation system tries to
implement the human translation rules into
computers processing rules. The idea of using
computers to translate human language is almost
as old as computer itself. However, Machine
Translation has been considered to be one of the
most difficult problems in Artificial
Intelligence.

• Project Details
• This project is an extension of a successful
  Master project, English to Vietnamese Machine
  Translation, by Van Hung Phan last semester.
  This project is implemented by the Example-based
  Machine Translation methodology concentrating on
  the quality of translation instead of more
  functionality. All issues in current Machine
  Translation technologies are considered to
  improve translations quality in this project
  (refer to previous section for project issues).
• Contribution of the Project
• Modify the Master project to translation from
  Chinese to English
• Consider linguistic issues between Chinese and
  English (language difference and structural
  ambiguity) in translation phase
• Add ontology into translation selection criteria
  (lexical ambiguity)
• Add more examples in database (multiword units)
• Improve database structure (processing speed and
  maintainability)
• Research on current trend and technology in
  Machine Translation
• Research on how to involve Artificial
  Intelligent in Machine Translation

• Current Methodologies in Machine Translation
• Direct Translation
• (eg. Georgetown system and SYSTRAN)
• Transfer Translation
• Interlingua Translation (eg. UNITRAN)
• Example-Based Translation
• and other less popular methodologies
• Altavistas BabelFish

Example-Based Methodology

• Current Issues in Machine Translation
• Lexical and Structural Ambiguity
• Language Difference
• Multiword Units
• Processing Speed
• Modularity and Maintainability

Output
System Architecture